diff --git a/CLAUDE.md b/CLAUDE.md
index 94cabf9..56cac5b 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -35,7 +35,7 @@ This is a Claude Skills collection repository containing:
 | 09 | ourdigital-backoffice | Business document creation | "create proposal", "견적서" |
 | 10 | ourdigital-skill-creator | Meta skill for creating skills | "create skill", "init skill" |
 
-### SEO Tools (11-30)
+### SEO Tools (11-32)
 
 | # | Skill | Purpose | Trigger |
 |---|-------|---------|---------|
@@ -47,22 +47,20 @@ This is a Claude Skills collection repository containing:
 | 16 | seo-schema-validator | Structured data validation | "validate schema", "JSON-LD" |
 | 17 | seo-schema-generator | Schema markup creation | "generate schema", "create JSON-LD" |
 | 18 | seo-local-audit | NAP, GBP, citations | "local SEO", "Google Business Profile" |
+| 19 | seo-keyword-strategy | Keyword expansion, intent, clustering, gaps | "keyword research", "keyword strategy" |
+| 20 | seo-serp-analysis | Google/Naver SERP features, competitor positions | "SERP analysis", "SERP features" |
+| 21 | seo-position-tracking | Rank monitoring, visibility scores, alerts | "rank tracking", "position monitoring" |
+| 22 | seo-link-building | Backlink audit, toxic links, link gaps | "backlink audit", "link building" |
+| 23 | seo-content-strategy | Content audit, decay, briefs, clusters | "content strategy", "content audit" |
+| 24 | seo-ecommerce | Product page audit, product schema | "e-commerce SEO", "product SEO" |
+| 25 | seo-kpi-framework | Unified KPIs, health scores, ROI | "SEO KPI", "SEO performance" |
+| 26 | seo-international | Hreflang, content parity, multi-language | "international SEO", "hreflang" |
+| 27 | seo-ai-visibility | AI search citations, brand radar, SOV | "AI visibility", "AI search" |
+| 28 | seo-knowledge-graph | Entity SEO, Knowledge Panel, PAA | "knowledge graph", "entity SEO" |
 | 29 | seo-gateway-architect | Gateway page strategy | "SEO strategy", "gateway pages" |
 | 30 | seo-gateway-builder | Gateway page content | "build gateway page" |
-
-**Future SEO Skills (19-28 reserved):**
-
-| # | Planned Skill | Status |
-|---|--------------|--------|
-| 19 | Keyword Strategy & Research | Planned |
-| 20 | SERP Analysis | Planned |
-| 21 | Position Tracking | Planned |
-| 22 | Link Building Diagnosis | Planned |
-| 23 | Content Strategy | Planned |
-| 24 | E-Commerce SEO | Planned |
-| 25 | SEO KPI & Performance Framework | Planned |
-| 26 | International SEO | Planned |
-| 27-28 | *(reserved)* | — |
+| 31 | seo-competitor-intel | Competitor profiling, benchmarking, threats | "competitor analysis", "competitive intel" |
+| 32 | seo-crawl-budget | Log analysis, bot profiling, crawl waste | "crawl budget", "log analysis" |
 
 ### GTM/GA Tools (60-69)
 
@@ -209,9 +207,20 @@ our-claude-skills/
 │   ├── 16-seo-schema-validator/
 │   ├── 17-seo-schema-generator/
 │   ├── 18-seo-local-audit/
-│   ├── 19-28 (reserved for future SEO skills)
+│   ├── 19-seo-keyword-strategy/
+│   ├── 20-seo-serp-analysis/
+│   ├── 21-seo-position-tracking/
+│   ├── 22-seo-link-building/
+│   ├── 23-seo-content-strategy/
+│   ├── 24-seo-ecommerce/
+│   ├── 25-seo-kpi-framework/
+│   ├── 26-seo-international/
+│   ├── 27-seo-ai-visibility/
+│   ├── 28-seo-knowledge-graph/
 │   ├── 29-seo-gateway-architect/
 │   ├── 30-seo-gateway-builder/
+│   ├── 31-seo-competitor-intel/
+│   ├── 32-seo-crawl-budget/
 │   │
 │   ├── 60-gtm-audit/
 │   ├── 61-gtm-manager/
diff --git a/custom-skills/11-seo-comprehensive-audit/code/CLAUDE.md b/custom-skills/11-seo-comprehensive-audit/code/CLAUDE.md
index 7e442cb..d03d20f 100644
--- a/custom-skills/11-seo-comprehensive-audit/code/CLAUDE.md
+++ b/custom-skills/11-seo-comprehensive-audit/code/CLAUDE.md
@@ -34,9 +34,38 @@ python scripts/seo_audit_orchestrator.py --url https://example.com --json
 | 2 | On-Page SEO | `13-seo-on-page-audit/code/scripts/page_analyzer.py` |
 | 3 | Core Web Vitals | `14-seo-core-web-vitals/code/scripts/pagespeed_client.py` |
 | 4 | Schema Validation | `16-seo-schema-validator/code/scripts/schema_validator.py` |
-| 5 | Local SEO | `18-seo-local-audit/` (prompt-driven) |
+| 5 | Local SEO | `18-seo-local-audit/` (prompt-driven — see Stage 5 notes below) |
 | 6 | Search Console | `15-seo-search-console/code/scripts/gsc_client.py` |
 
+## Stage 5: Local SEO — Key Requirements
+
+Stage 5 is prompt-driven and requires **Business Identity extraction as a mandatory first step**:
+1. Extract Korean name, English name, address, phone from website JSON-LD schema markup (`Organization`/`Hospital`/`LocalBusiness`)
+2. Check website footer, contact page, and schema `sameAs` for GBP, Naver Place, and Kakao Map URLs
+3. Use layered search fallback if listing URLs are not found on the website
+4. Follow `18-seo-local-audit/code/CLAUDE.md` for the full workflow
+5. **Korean market priorities**: GBP and Naver Smart Place are both Critical; Kakao Map is High; US-centric directories (Yelp, Yellow Pages) are Low
+6. **Important**: GBP and Naver Map are JS-rendered. Report unfound listings as "not discoverable via web search" — not "does not exist"
+
+## Extended SEO Skills Pipeline
+
+Beyond the 6 core audit stages, additional specialized skills are available for deeper analysis:
+
+| Skill | Audit ID | Purpose | Command |
+|-------|----------|---------|---------|
+| 19 - Keyword Strategy | KW | Seed expansion, intent classification, keyword gaps | `/seo-keyword-strategy` |
+| 20 - SERP Analysis | SERP | Google/Naver SERP features, competitor positions | `/seo-serp-analysis` |
+| 21 - Position Tracking | RANK | Rank monitoring, visibility scores, alerts | `/seo-position-tracking` |
+| 22 - Link Building | LINK | Backlink audit, toxic links, link gaps | `/seo-link-building` |
+| 23 - Content Strategy | CONTENT | Content audit, decay detection, briefs | `/seo-content-strategy` |
+| 24 - E-Commerce SEO | ECOM | Product page audit, product schema | `/seo-ecommerce` |
+| 25 - SEO KPI Framework | KPI | Unified KPIs, health scores, ROI | `/seo-kpi-framework` |
+| 26 - International SEO | INTL | Hreflang validation, content parity | `/seo-international` |
+| 27 - AI Visibility | AI | AI search citations, brand radar, SOV | `/seo-ai-visibility` |
+| 28 - Knowledge Graph | KG | Entity SEO, Knowledge Panel, PAA | `/seo-knowledge-graph` |
+| 31 - Competitor Intel | COMP | Competitor profiling, benchmarking | `/seo-competitor-intel` |
+| 32 - Crawl Budget | CRAWL | Log analysis, bot profiling, waste | `/seo-crawl-budget` |
+
 ## Health Score Weights
 
 | Category | Weight |
diff --git a/custom-skills/11-seo-comprehensive-audit/code/commands/seo-comprehensive-audit.md b/custom-skills/11-seo-comprehensive-audit/code/commands/seo-comprehensive-audit.md
index 00cbfab..ff09298 100644
--- a/custom-skills/11-seo-comprehensive-audit/code/commands/seo-comprehensive-audit.md
+++ b/custom-skills/11-seo-comprehensive-audit/code/commands/seo-comprehensive-audit.md
@@ -62,10 +62,37 @@ python "$SKILLS/14-seo-core-web-vitals/code/scripts/pagespeed_client.py" --url $
 # Stage 4: Schema Validation
 python "$SKILLS/16-seo-schema-validator/code/scripts/schema_validator.py" --url $URL --json
 
-# Stage 5: Local SEO (prompt-driven, use WebFetch + WebSearch)
+# Stage 5: Local SEO (see detailed instructions below)
 # Stage 6: Search Console (requires GSC API credentials)
 ```
 
+### Stage 5: Local SEO — Detailed Instructions
+
+Stage 5 is prompt-driven (no script). Follow this sequence:
+
+1. **Extract Business Identity from website (MANDATORY FIRST)**
+   - WebFetch the homepage and parse JSON-LD `<script type="application/ld+json">` tags
+   - Extract from `Organization`, `Hospital`, or `LocalBusiness` schema: Korean name, English name, address, telephone
+   - Check `sameAs` array for GBP, Naver Place, Kakao Map URLs
+
+2. **Check website for listing links**
+   - Scrape footer, contact page, about page for links matching:
+     - GBP: `maps.app.goo.gl/*`, `google.com/maps/place/*`, `g.page/*`
+     - Naver Place: `naver.me/*`, `map.naver.com/*/place/*`, `m.place.naver.com/*`
+     - Kakao Map: `place.map.kakao.com/*`, `kko.to/*`
+   - Check embedded iframes for Google Maps Place IDs or Naver Map embeds
+
+3. **Layered search fallback (if links not found on website)**
+   - GBP: Search `"[Korean Name]" "[district]" Google Maps`, then `"[phone]" site:google.com/maps`
+   - Naver: Search `"[Korean Name]" site:map.naver.com`, then `"[Korean Name]" 네이버 지도 [district]`
+   - Kakao: Search `"[Korean Name]" site:place.map.kakao.com`
+
+4. **Follow `18-seo-local-audit/code/CLAUDE.md` workflow** for the full audit (Steps 2-7)
+
+5. **Important language**: Distinguish **"not discoverable via web search"** from **"does not exist."** GBP and Naver Map are JS-rendered; WebFetch cannot extract their listing data. Absence in search results does not confirm absence of the listing.
+
+6. **Korean market priorities**: GBP and Naver Smart Place are both Critical. Kakao Map is High. US-centric directories (Yelp, Yellow Pages) are Low priority for Korean businesses.
+
 ## Health Score (Weighted 0-100)
 
 | Category | Weight |
diff --git a/custom-skills/18-seo-local-audit/code/CLAUDE.md b/custom-skills/18-seo-local-audit/code/CLAUDE.md
index 2395b94..e621982 100644
--- a/custom-skills/18-seo-local-audit/code/CLAUDE.md
+++ b/custom-skills/18-seo-local-audit/code/CLAUDE.md
@@ -2,109 +2,253 @@
 
 ## Overview
 
-Local SEO auditor for businesses with physical locations: NAP consistency, Google Business Profile optimization, local citations, and LocalBusiness schema validation.
-
-## Quick Start
-
-This skill primarily uses MCP tools (Firecrawl, Perplexity) for data collection. Scripts are helpers for validation.
-
-```bash
-# NAP consistency check (manual data input)
-python scripts/nap_checker.py --business "Business Name" --address "123 Main St" --phone "555-1234"
-
-# LocalBusiness schema validation
-python scripts/local_schema_validator.py --url https://example.com
-```
-
-## Audit Components
-
-### 1. NAP Consistency
-**Name, Address, Phone** consistency across:
-- Website (header, footer, contact page)
-- Google Business Profile
-- Local directories (Yelp, Yellow Pages, etc.)
-- Social media profiles
-
-### 2. Google Business Profile (GBP)
-Optimization checklist:
-- [ ] Business name matches website
-- [ ] Address is complete and accurate
-- [ ] Phone number is local
-- [ ] Business hours are current
-- [ ] Categories are appropriate
-- [ ] Photos uploaded (exterior, interior, products)
-- [ ] Posts are recent (within 7 days)
-- [ ] Reviews are responded to
-
-### 3. Local Citations
-Priority directories to check:
-- Google Business Profile
-- Apple Maps
-- Bing Places
-- Yelp
-- Facebook Business
-- Industry-specific directories
-
-### 4. LocalBusiness Schema
-Required properties:
-- @type (LocalBusiness or subtype)
-- name
-- address (PostalAddress)
-- telephone
-- openingHours
+Local SEO auditor for Korean-market businesses with physical locations. Covers business identity extraction, GBP optimization, Naver Smart Place, Kakao Map, NAP consistency, local citations, and LocalBusiness schema validation.
 
 ## Workflow
 
+### Step 0: Business Identity (MANDATORY FIRST STEP)
+
+Before any audit work, establish the official business identity.
+
+**Sources (in priority order):**
+1. Website schema markup (JSON-LD `Organization`, `Hospital`, `LocalBusiness`) — the `name` field is authoritative
+2. Contact page / About page
+3. Footer (address, phone, social links)
+4. User-provided information (known GBP URL, Naver Place URL, etc.)
+
+**Data to collect:**
+
+| Field | Example |
+|-------|---------|
+| Official name (Korean) | 제이미성형외과의원 |
+| Official name (English) | Jamie Plastic Surgery Clinic |
+| Brand/display name | Jamie Clinic |
+| Website URL | https://www.jamie.clinic |
+| Address (Korean) | 서울특별시 강남구 ... |
+| Phone | 02-XXX-XXXX |
+| Known GBP URL | (if available) |
+| Known Naver Place URL | (if available) |
+| Known Kakao Map URL | (if available) |
+
+**How to extract:**
 ```
-1. Collect NAP from client
-2. Scrape website for NAP mentions
-3. Search citations using Perplexity
-4. Check GBP data (manual or API)
-5. Validate LocalBusiness schema
-6. Generate consistency report
+WebFetch homepage → parse JSON-LD script tags → extract name, address, telephone, sameAs
+WebFetch /contact or /about → extract NAP from page content
+Check footer for social links, map embeds, place listing URLs
 ```
 
+Look specifically for these URL patterns in `sameAs`, footer links, or embedded iframes:
+- GBP: `maps.app.goo.gl/*`, `google.com/maps/place/*`, `g.page/*`
+- Naver Place: `naver.me/*`, `map.naver.com/*/place/*`, `m.place.naver.com/*`
+- Kakao Map: `place.map.kakao.com/*`, `kko.to/*`
+
+### Step 1: Website NAP Extraction
+
+Scrape header, footer, contact page, about page for NAP mentions. Cross-reference with schema markup. Establish the **canonical NAP** baseline (the single source of truth for this audit).
+
+### Step 2: GBP Verification & Audit
+
+**Layered discovery (try in order, stop when found):**
+1. Use provided GBP URL (from Step 0 or user input)
+2. Check website for GBP link (footer, contact page, schema `sameAs`, embedded Google Maps iframe with Place ID)
+3. WebSearch: `"[Korean Name]" "[City/District]" Google Maps`
+4. WebSearch: `"[English Name]" Google Maps [City]`
+5. WebSearch: `"[exact phone number]" site:google.com/maps`
+
+**Important**: Google Maps is JS-rendered — WebFetch cannot extract business data from the listing page itself. Use WebSearch to find the listing URL, then verify details via search result snippets.
+
+**If found — audit checklist (score /10):**
+- [ ] Business name matches canonical NAP
+- [ ] Address is complete and accurate
+- [ ] Phone number matches
+- [ ] Business hours are current
+- [ ] Primary + secondary categories are appropriate
+- [ ] Business description is complete
+- [ ] 10+ photos uploaded (exterior, interior, products/services)
+- [ ] Posts are recent (within 7 days)
+- [ ] Reviews are responded to
+- [ ] Q&A section is active
+
+**If NOT found after all attempts:** Report as **"not discoverable via web search"** — this is distinct from "does not exist." The listing may exist but be unfindable through non-JS search methods.
+
+### Step 3: Naver Smart Place Verification & Audit
+
+**Layered discovery (try in order, stop when found):**
+1. Use provided Naver Place URL (from Step 0 or user input)
+2. Check website for Naver Place link (footer, contact page, schema `sameAs`, `naver.me/*` or `map.naver.com/*/place/*` patterns)
+3. WebSearch: `"[Korean Name]" site:map.naver.com`
+4. WebSearch: `"[Korean Name]" 네이버 지도 [district]`
+5. WebSearch: `"[Korean Name]" 네이버 스마트플레이스`
+6. WebSearch: `"[exact phone number]" site:map.naver.com`
+
+**Important**: Naver Map is JS-rendered — WebFetch cannot extract data from the listing page. Use WebSearch for discovery, verify via search result snippets.
+
+**If found — audit checklist (score /10):**
+- [ ] Business name matches canonical NAP
+- [ ] Address is complete and accurate
+- [ ] Phone number matches
+- [ ] Business hours are current
+- [ ] Place is "claimed" (owner-managed / 업주 등록)
+- [ ] Keywords/tags are set
+- [ ] Booking/reservation link present
+- [ ] Recent blog reviews linked
+- [ ] Photos uploaded and current
+- [ ] Menu/service/price information present
+
+**If NOT found after all attempts:** Report as **"not discoverable via web search"** (not "does not exist" or "not registered").
+
+### Step 4: Kakao Map Verification
+
+**Discovery:**
+1. Use provided Kakao Map URL (from Step 0)
+2. Check website for Kakao Map link (`place.map.kakao.com/*`, `kko.to/*`)
+3. WebSearch: `"[Korean Name]" site:place.map.kakao.com`
+4. WebSearch: `"[Korean Name]" 카카오맵 [district]`
+
+**If found:** Verify NAP consistency against canonical NAP.
+
+### Step 5: Citation Discovery
+
+**Korean market platform priorities:**
+
+| Platform | Priority | Market |
+|----------|----------|--------|
+| Google Business Profile | Critical | Global |
+| Naver Smart Place (네이버 스마트플레이스) | Critical | Korea |
+| Kakao Map (카카오맵) | High | Korea |
+| Industry-specific directories | High | Varies |
+| Apple Maps | Medium | Global |
+| Bing Places | Low | Global |
+
+**Korean medical/cosmetic industry directories:**
+- 강남언니 (Gangnam Unni)
+- 바비톡 (Babitalk)
+- 성예사 (Sungyesa)
+- 굿닥 (Goodoc)
+- 똑닥 (Ddocdoc)
+- 모두닥 (Modoodoc)
+- 하이닥 (HiDoc)
+
+**Discovery methods:**
+- Phone number search across platforms
+- Korean business name + district search
+- English business name search
+- Address search
+
+### Step 6: NAP Consistency Report
+
+Cross-reference all discovered sources against the canonical NAP from Step 1.
+
+**Common inconsistency points to check:**
+- Building/landmark names (e.g., "EHL빌딩" vs "엔와이빌딩") — the authoritative source is the **business registration certificate** (사업자등록증), not the website alone
+- Phone format variations (02-XXX-XXXX vs +82-2-XXX-XXXX vs 02XXXXXXX)
+- Address format (road-name vs lot-number / 도로명 vs 지번)
+- Korean vs English name spelling variations
+- Suite/floor number omissions
+
+### Step 7: LocalBusiness Schema Validation
+
+Validate JSON-LD completeness:
+- @type (LocalBusiness, Hospital, or appropriate subtype)
+- name (Korean and/or English)
+- address (PostalAddress with Korean format)
+- telephone
+- openingHours / openingHoursSpecification
+- geo (GeoCoordinates — latitude, longitude)
+- sameAs (should include GBP, Naver Place, Kakao Map, social profiles)
+- url
+- image
+
+Use schema generator skill (17) for creating/fixing markup.
+
+## Scoring
+
+| Component | Weight | Max Score |
+|-----------|--------|-----------|
+| Business Identity completeness | 5% | /10 |
+| NAP Consistency | 20% | /10 |
+| GBP Optimization | 20% | /10 |
+| Naver Smart Place | 20% | /10 |
+| Kakao Map presence | 10% | /10 |
+| Citations (directories) | 10% | /10 |
+| LocalBusiness Schema | 15% | /10 |
+
+**Overall Local SEO Score** = weighted average, normalized to /100.
+
 ## Output Format
 
 ```markdown
 ## Local SEO Audit: [Business Name]
+**Date**: YYYY-MM-DD
+**Website**: [URL]
 
-### NAP Consistency Score: X/10
+### Business Identity
+| Field | Value |
+|-------|-------|
+| Korean Name | ... |
+| English Name | ... |
+| Brand Name | ... |
+| Address | ... |
+| Phone | ... |
 
+### NAP Consistency: X/10
 | Source | Name | Address | Phone | Status |
 |--------|------|---------|-------|--------|
-| Website | ✓ | ✓ | ✓ | Match |
-| GBP | ✓ | ✗ | ✓ | Mismatch |
+| Website | OK/Issue | OK/Issue | OK/Issue | Match/Mismatch |
+| GBP | OK/Issue | OK/Issue | OK/Issue | Match/Mismatch |
+| Naver Place | OK/Issue | OK/Issue | OK/Issue | Match/Mismatch |
+| Kakao Map | OK/Issue | OK/Issue | OK/Issue | Match/Mismatch |
 
 ### GBP Optimization: X/10
-- [ ] Issue 1
-- [x] Completed item
+- [x] Completed items
+- [ ] Missing items
+**GBP URL**: [URL or "not discoverable"]
 
-### Citation Audit
-- Found: X citations
-- Consistent: X
-- Needs update: X
+### Naver Smart Place: X/10
+- [x] Completed items
+- [ ] Missing items
+**Naver Place URL**: [URL or "not discoverable"]
 
-### Recommendations
-1. Fix address mismatch on GBP
-2. Add LocalBusiness schema
+### Kakao Map: X/10
+**Status**: Found/Not discoverable
+**Kakao Map URL**: [URL or "not discoverable"]
+
+### Citations: X/10
+| Platform | Found | NAP Match |
+|----------|-------|-----------|
+| 강남언니 | Yes/No | OK/Issue |
+| ... | | |
+
+### LocalBusiness Schema: X/10
+- Present: Yes/No
+- Valid: Yes/No
+- Missing fields: [list]
+
+### Overall Score: XX/100 (Grade)
+### Priority Actions
+1. [Highest impact recommendation]
+2. ...
 ```
 
 ## Common Issues
 
 | Issue | Impact | Fix |
 |-------|--------|-----|
-| NAP inconsistency | High | Update all directories |
-| Missing GBP categories | Medium | Add relevant categories |
-| No LocalBusiness schema | Medium | Add JSON-LD markup |
-| Outdated business hours | Medium | Update GBP hours |
-| No review responses | Low | Respond to all reviews |
+| NAP inconsistency | High | Update all directories to match canonical NAP |
+| Missing Naver Smart Place | Critical | Register and claim via smartplace.naver.com |
+| Unclaimed Naver Place | High | Claim ownership via 네이버 스마트플레이스 |
+| Missing GBP listing | Critical | Create via business.google.com |
+| Building name mismatch | Medium | Align to business registration certificate |
+| No LocalBusiness schema | Medium | Add JSON-LD markup with sameAs links |
+| Missing GeoCoordinates | Medium | Add lat/lng to schema |
+| No sameAs in schema | Medium | Add GBP, Naver, Kakao, social URLs |
 
 ## Notes
 
-- GBP API requires enterprise approval (use manual audit)
-- Citation discovery limited to public data
-- Use schema generator skill (14) for creating LocalBusiness markup
+- GBP and Naver Map are JS-rendered — WebFetch cannot extract listing data directly. Always use WebSearch for discovery.
+- "Not discoverable via web search" != "does not exist." Always use this precise language.
+- For Korean businesses, Naver Smart Place is as important as GBP (often more so for domestic traffic).
+- Citation discovery is limited to publicly searchable data.
 
 ## Notion Output (Required)
 
@@ -123,20 +267,13 @@ Required properties:
 |----------|------|-------------|
 | Issue | Title | Report title (Korean + date) |
 | Site | URL | Audited website URL |
-| Category | Select | Technical SEO, On-page SEO, Performance, Schema/Structured Data, Sitemap, Robots.txt, Content, Local SEO |
+| Category | Select | Local SEO |
 | Priority | Select | Critical, High, Medium, Low |
 | Found Date | Date | Audit date (YYYY-MM-DD) |
-| Audit ID | Rich Text | Format: [TYPE]-YYYYMMDD-NNN |
+| Audit ID | Rich Text | Format: LOCAL-YYYYMMDD-NNN |
 
 ### Language Guidelines
 
 - Report content in Korean (한국어)
-- Keep technical English terms as-is (e.g., SEO Audit, Core Web Vitals, Schema Markup)
+- Keep technical English terms as-is (e.g., SEO Audit, GBP, NAP, Schema Markup)
 - URLs and code remain unchanged
-
-### Example MCP Call
-
-```bash
-mcp-cli call notion/API-post-page '{"parent": {"database_id": "2c8581e5-8a1e-8035-880b-e38cefc2f3ef"}, "properties": {...}}'
-```
-
diff --git a/custom-skills/18-seo-local-audit/desktop/SKILL.md b/custom-skills/18-seo-local-audit/desktop/SKILL.md
index 8b8efd0..787599c 100644
--- a/custom-skills/18-seo-local-audit/desktop/SKILL.md
+++ b/custom-skills/18-seo-local-audit/desktop/SKILL.md
@@ -1,125 +1,239 @@
 ---
 name: seo-local-audit
 description: |
-  Local business SEO auditor for NAP consistency, Google Business Profile, and citations.
-  Triggers: local SEO, NAP audit, Google Business Profile, GBP optimization, local citations.
+  Local business SEO auditor for Korean-market businesses. Covers business identity extraction,
+  NAP consistency, Google Business Profile, Naver Smart Place, Kakao Map, local citations,
+  and LocalBusiness schema validation.
+  Triggers: local SEO, NAP audit, Google Business Profile, GBP optimization, local citations,
+  네이버 스마트플레이스, 카카오맵, 로컬 SEO.
 ---
 
 # SEO Local Audit
 
 ## Purpose
 
-Audit local business SEO: NAP (Name, Address, Phone) consistency, Google Business Profile optimization, local citations, and LocalBusiness schema markup.
+Audit local business SEO for Korean-market businesses: business identity extraction, NAP consistency, GBP optimization, Naver Smart Place, Kakao Map, local citations, and LocalBusiness schema markup.
 
 ## Core Capabilities
 
-1. **NAP Consistency** - Cross-platform verification
-2. **GBP Optimization** - Profile completeness check
-3. **Citation Audit** - Directory presence
-4. **Schema Validation** - LocalBusiness markup
+1. **Business Identity** - Extract official names, address, phone from website schema/content
+2. **NAP Consistency** - Cross-platform verification against canonical NAP
+3. **GBP Optimization** - Layered discovery + profile completeness audit
+4. **Naver Smart Place** - Layered discovery + listing completeness audit
+5. **Kakao Map** - Presence verification + NAP check
+6. **Citation Audit** - Korean-first directory presence
+7. **Schema Validation** - LocalBusiness JSON-LD markup
 
 ## MCP Tool Usage
 
 ```
-mcp__firecrawl__scrape: Extract NAP from website
-mcp__perplexity__search: Find citations and directories
+mcp__firecrawl__scrape: Extract NAP and schema from website
+mcp__perplexity__search: Find citations, GBP, Naver Place listings
 mcp__notion__create-page: Save audit findings
 ```
 
 ## Workflow
 
-### 1. Gather Business Info
-Collect from client:
-- Business name (exact)
-- Full address
-- Phone number (local preferred)
-- Website URL
-- GBP listing URL
+### Step 0: Business Identity (MANDATORY FIRST STEP)
 
-### 2. Website NAP Check
-Scrape website for NAP mentions:
-- Header/footer
-- Contact page
-- About page
-- Schema markup
+Before any audit, establish the official business identity.
 
-### 3. Citation Discovery
-Search for business mentions:
-- "[Business Name] [City]"
-- Phone number search
-- Address search
+**Sources (in priority order):**
+1. Website schema markup (JSON-LD `Organization`, `Hospital`, `LocalBusiness`) — `name` field is authoritative
+2. Contact page / About page
+3. Footer (address, phone, social links)
+4. User-provided information
 
-### 4. GBP Review
-Manual checklist:
-- Profile completeness
-- Category accuracy
-- Photo presence
-- Review responses
-- Post recency
+**Data to collect:**
 
-### 5. Schema Check
-Validate LocalBusiness markup presence and accuracy.
+| Field | Example |
+|-------|---------|
+| Official name (Korean) | 제이미성형외과의원 |
+| Official name (English) | Jamie Plastic Surgery Clinic |
+| Brand/display name | Jamie Clinic |
+| Website URL | https://www.jamie.clinic |
+| Address (Korean) | 서울특별시 강남구 ... |
+| Phone | 02-XXX-XXXX |
+| Known GBP URL | (if available) |
+| Known Naver Place URL | (if available) |
+| Known Kakao Map URL | (if available) |
 
-## GBP Optimization Checklist
+Look for these URL patterns in `sameAs`, footer links, or embedded iframes:
+- GBP: `maps.app.goo.gl/*`, `google.com/maps/place/*`, `g.page/*`
+- Naver Place: `naver.me/*`, `map.naver.com/*/place/*`, `m.place.naver.com/*`
+- Kakao Map: `place.map.kakao.com/*`, `kko.to/*`
 
-- [ ] Business name matches website
-- [ ] Complete address with suite/unit
-- [ ] Local phone number (not toll-free)
-- [ ] Accurate business hours
-- [ ] Primary + secondary categories set
+### Step 1: Website NAP Extraction
+
+Scrape header, footer, contact page, about page. Cross-reference with schema markup. Establish the **canonical NAP** baseline.
+
+### Step 2: GBP Verification & Audit
+
+**Layered discovery (try in order, stop when found):**
+1. Use provided GBP URL (from Step 0 or user input)
+2. Check website for GBP link (footer, contact, schema `sameAs`, embedded Google Maps iframe)
+3. Search: `"[Korean Name]" "[City/District]" Google Maps`
+4. Search: `"[English Name]" Google Maps [City]`
+5. Search: `"[exact phone number]" site:google.com/maps`
+
+**Important**: Google Maps is JS-rendered — scraping tools cannot extract business data. Use search for discovery, verify via search result snippets.
+
+**If found — audit checklist (score /10):**
+- [ ] Business name matches canonical NAP
+- [ ] Address is complete and accurate
+- [ ] Phone number matches
+- [ ] Business hours are current
+- [ ] Primary + secondary categories appropriate
 - [ ] Business description complete
 - [ ] 10+ photos uploaded
-- [ ] Recent post (within 7 days)
-- [ ] Reviews responded to
+- [ ] Posts are recent (within 7 days)
+- [ ] Reviews are responded to
+- [ ] Q&A section is active
 
-## Citation Priority
+**If NOT found:** Report as **"not discoverable via web search"** (distinct from "does not exist").
 
-| Platform | Priority |
-|----------|----------|
-| Google Business Profile | Critical |
-| Apple Maps | High |
-| Bing Places | High |
-| Yelp | High |
-| Facebook | Medium |
-| Industry directories | Medium |
+### Step 3: Naver Smart Place Verification & Audit
+
+**Layered discovery (try in order, stop when found):**
+1. Use provided Naver Place URL (from Step 0 or user input)
+2. Check website for Naver Place link (footer, contact, schema `sameAs`)
+3. Search: `"[Korean Name]" site:map.naver.com`
+4. Search: `"[Korean Name]" 네이버 지도 [district]`
+5. Search: `"[Korean Name]" 네이버 스마트플레이스`
+6. Search: `"[exact phone number]" site:map.naver.com`
+
+**Important**: Naver Map is JS-rendered — scraping tools cannot extract data. Use search for discovery, verify via snippets.
+
+**If found — audit checklist (score /10):**
+- [ ] Business name matches canonical NAP
+- [ ] Address is complete and accurate
+- [ ] Phone number matches
+- [ ] Business hours are current
+- [ ] Place is "claimed" (owner-managed / 업주 등록)
+- [ ] Keywords/tags are set
+- [ ] Booking/reservation link present
+- [ ] Recent blog reviews linked
+- [ ] Photos uploaded and current
+- [ ] Menu/service/price information present
+
+**If NOT found:** Report as **"not discoverable via web search"** (not "does not exist" or "not registered").
+
+### Step 4: Kakao Map Verification
+
+**Discovery:**
+1. Use provided Kakao Map URL (from Step 0)
+2. Check website for Kakao Map link (`place.map.kakao.com/*`, `kko.to/*`)
+3. Search: `"[Korean Name]" site:place.map.kakao.com`
+4. Search: `"[Korean Name]" 카카오맵 [district]`
+
+**If found:** Verify NAP consistency against canonical NAP.
+
+### Step 5: Citation Discovery
+
+**Korean market platform priorities:**
+
+| Platform | Priority | Market |
+|----------|----------|--------|
+| Google Business Profile | Critical | Global |
+| Naver Smart Place (네이버 스마트플레이스) | Critical | Korea |
+| Kakao Map (카카오맵) | High | Korea |
+| Industry-specific directories | High | Varies |
+| Apple Maps | Medium | Global |
+| Bing Places | Low | Global |
+
+**Korean medical/cosmetic industry directories:**
+- 강남언니 (Gangnam Unni)
+- 바비톡 (Babitalk)
+- 성예사 (Sungyesa)
+- 굿닥 (Goodoc)
+- 똑닥 (Ddocdoc)
+- 모두닥 (Modoodoc)
+- 하이닥 (HiDoc)
+
+### Step 6: NAP Consistency Report
+
+Cross-reference all sources against canonical NAP.
+
+**Common inconsistency points:**
+- Building/landmark names — authoritative source is the **business registration certificate** (사업자등록증)
+- Phone format variations (02-XXX-XXXX vs +82-2-XXX-XXXX)
+- Address format (road-name vs lot-number / 도로명 vs 지번)
+- Korean vs English name spelling variations
+- Suite/floor number omissions
+
+### Step 7: LocalBusiness Schema Validation
+
+Validate JSON-LD completeness: @type, name, address, telephone, openingHours, geo (GeoCoordinates), sameAs (GBP, Naver, Kakao, social), url, image.
+
+## Scoring
+
+| Component | Weight | Max Score |
+|-----------|--------|-----------|
+| Business Identity completeness | 5% | /10 |
+| NAP Consistency | 20% | /10 |
+| GBP Optimization | 20% | /10 |
+| Naver Smart Place | 20% | /10 |
+| Kakao Map presence | 10% | /10 |
+| Citations (directories) | 10% | /10 |
+| LocalBusiness Schema | 15% | /10 |
+
+**Overall Local SEO Score** = weighted average, normalized to /100.
 
 ## Output Format
 
 ```markdown
 ## Local SEO Audit: [Business]
 
+### Business Identity
+| Field | Value |
+|-------|-------|
+| Korean Name | ... |
+| English Name | ... |
+| Address | ... |
+| Phone | ... |
+
 ### NAP Consistency: X/10
-| Source | Name | Address | Phone |
-|--------|------|---------|-------|
-| Website | ✓/✗ | ✓/✗ | ✓/✗ |
-| GBP | ✓/✗ | ✓/✗ | ✓/✗ |
+| Source | Name | Address | Phone | Status |
+|--------|------|---------|-------|--------|
+| Website | OK/Issue | OK/Issue | OK/Issue | Match/Mismatch |
+| GBP | OK/Issue | OK/Issue | OK/Issue | Match/Mismatch |
+| Naver Place | OK/Issue | OK/Issue | OK/Issue | Match/Mismatch |
+| Kakao Map | OK/Issue | OK/Issue | OK/Issue | Match/Mismatch |
 
 ### GBP Score: X/10
 [Checklist results]
 
-### Citations Found: X
-- Consistent: X
-- Inconsistent: X
+### Naver Smart Place: X/10
+[Checklist results]
 
-### LocalBusiness Schema
+### Kakao Map: X/10
+[Status + NAP check]
+
+### Citations: X/10
+| Platform | Found | NAP Match |
+|----------|-------|-----------|
+| ... | | |
+
+### LocalBusiness Schema: X/10
 - Present: Yes/No
 - Valid: Yes/No
+- Missing fields: [list]
 
+### Overall Score: XX/100 (Grade)
 ### Priority Actions
-1. [Fix recommendations]
+1. [Recommendations]
 ```
 
-## Limitations
+## Notes
 
-- GBP data requires manual access
-- Citation discovery limited to searchable sources
-- Cannot update external directories
+- GBP and Naver Map are JS-rendered — scraping tools cannot extract listing data. Always use search for discovery.
+- "Not discoverable via web search" != "does not exist." Always use this precise language.
+- For Korean businesses, Naver Smart Place is as important as GBP (often more so for domestic traffic).
 
 ## Notion Output (Required)
 
 All audit reports MUST be saved to OurDigital SEO Audit Log:
 - **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
-- **Properties**: Issue (title), Site (url), Category, Priority, Found Date, Audit ID
+- **Properties**: Issue (title), Site (url), Category (Local SEO), Priority, Found Date, Audit ID
 - **Language**: Korean with English technical terms
-- **Audit ID Format**: [TYPE]-YYYYMMDD-NNN
-
+- **Audit ID Format**: LOCAL-YYYYMMDD-NNN
diff --git a/custom-skills/19-seo-keyword-strategy/code/CLAUDE.md b/custom-skills/19-seo-keyword-strategy/code/CLAUDE.md
new file mode 100644
index 0000000..a9caa3f
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/code/CLAUDE.md
@@ -0,0 +1,139 @@
+# CLAUDE.md
+
+## Overview
+
+Keyword strategy and research tool for SEO campaigns. Expands seed keywords via Ahrefs APIs, classifies search intent, clusters topics, performs competitor keyword gap analysis, and supports Korean market keyword discovery including Naver autocomplete.
+
+## Quick Start
+
+```bash
+# Install dependencies
+pip install -r scripts/requirements.txt
+
+# Keyword research from seed keyword
+python scripts/keyword_researcher.py --keyword "치과 임플란트" --country kr --json
+
+# Keyword gap analysis vs competitor
+python scripts/keyword_gap_analyzer.py --target https://example.com --competitor https://competitor.com --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `keyword_researcher.py` | Expand seed keywords, classify intent, cluster topics | Keyword list with volume, KD, intent, clusters |
+| `keyword_gap_analyzer.py` | Find competitor keyword gaps | Gap keywords with opportunity scores |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## Keyword Researcher
+
+```bash
+# Basic expansion
+python scripts/keyword_researcher.py --keyword "dental implant" --json
+
+# Korean market with suffix expansion
+python scripts/keyword_researcher.py --keyword "치과 임플란트" --country kr --korean-suffixes --json
+
+# With volume-by-country comparison
+python scripts/keyword_researcher.py --keyword "dental implant" --country kr --compare-global --json
+
+# Output to file
+python scripts/keyword_researcher.py --keyword "치과 임플란트" --country kr --output report.json
+```
+
+**Capabilities**:
+- Seed keyword expansion (matching terms, related terms, search suggestions)
+- Korean suffix expansion (추천, 가격, 후기, 잘하는곳, 부작용, 전후)
+- Search intent classification (informational/navigational/commercial/transactional)
+- Keyword clustering into topic groups
+- Volume-by-country comparison (Korea vs global)
+- Keyword difficulty scoring
+
+## Keyword Gap Analyzer
+
+```bash
+# Find gaps vs one competitor
+python scripts/keyword_gap_analyzer.py --target https://example.com --competitor https://competitor.com --json
+
+# Multiple competitors
+python scripts/keyword_gap_analyzer.py --target https://example.com --competitor https://comp1.com --competitor https://comp2.com --json
+
+# Filter by minimum volume
+python scripts/keyword_gap_analyzer.py --target https://example.com --competitor https://competitor.com --min-volume 100 --json
+```
+
+**Capabilities**:
+- Identify keywords competitors rank for but target doesn't
+- Opportunity scoring based on volume, KD, and competitor positions
+- Segment gaps by intent type
+- Prioritize low-KD high-volume opportunities
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `keywords-explorer-overview` | Get keyword metrics (volume, KD, CPC) |
+| `keywords-explorer-matching-terms` | Find matching keyword variations |
+| `keywords-explorer-related-terms` | Discover semantically related keywords |
+| `keywords-explorer-search-suggestions` | Get autocomplete suggestions |
+| `keywords-explorer-volume-by-country` | Compare volume across countries |
+| `keywords-explorer-volume-history` | Track volume trends over time |
+| `site-explorer-organic-keywords` | Get competitor keyword rankings |
+
+## Output Format
+
+All scripts support `--json` flag for structured output:
+
+```json
+{
+  "seed_keyword": "치과 임플란트",
+  "country": "kr",
+  "total_keywords": 150,
+  "clusters": [
+    {
+      "topic": "임플란트 가격",
+      "keywords": [...],
+      "total_volume": 12000
+    }
+  ],
+  "keywords": [
+    {
+      "keyword": "치과 임플란트 가격",
+      "volume": 5400,
+      "kd": 32,
+      "cpc": 2.5,
+      "intent": "commercial",
+      "cluster": "임플란트 가격"
+    }
+  ],
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Audited website URL |
+| Category | Select | Keyword Research |
+| Priority | Select | Based on opportunity score |
+| Found Date | Date | Research date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: KW-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., Keyword Difficulty, Search Volume, CPC)
+- URLs and code remain unchanged
diff --git a/custom-skills/19-seo-keyword-strategy/code/scripts/base_client.py b/custom-skills/19-seo-keyword-strategy/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/19-seo-keyword-strategy/code/scripts/keyword_gap_analyzer.py b/custom-skills/19-seo-keyword-strategy/code/scripts/keyword_gap_analyzer.py
new file mode 100644
index 0000000..8a01ad2
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/code/scripts/keyword_gap_analyzer.py
@@ -0,0 +1,584 @@
+"""
+Keyword Gap Analyzer - Competitor keyword gap analysis with opportunity scoring
+===============================================================================
+Purpose: Identify keywords competitors rank for but target site doesn't,
+         score opportunities, and prioritize by volume/difficulty ratio.
+Python: 3.10+
+"""
+
+import argparse
+import json
+import logging
+import re
+import subprocess
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from typing import Optional
+from urllib.parse import urlparse
+
+# ---------------------------------------------------------------------------
+# Logging
+# ---------------------------------------------------------------------------
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+logger = logging.getLogger("keyword_gap_analyzer")
+
+# ---------------------------------------------------------------------------
+# Intent classification patterns (shared with keyword_researcher)
+# ---------------------------------------------------------------------------
+INTENT_PATTERNS: dict[str, list[str]] = {
+    "transactional": [
+        r"구매|구입|주문|buy|order|purchase|shop|deal|discount|coupon|할인|쿠폰",
+        r"예약|booking|reserve|sign\s?up|register|등록|신청",
+    ],
+    "commercial": [
+        r"가격|비용|얼마|price|cost|pricing|fee|요금",
+        r"추천|best|top\s?\d|review|비교|compare|vs|versus|후기|리뷰|평점|평가",
+        r"잘하는곳|잘하는|맛집|업체|병원|추천\s?병원",
+    ],
+    "navigational": [
+        r"^(www\.|http|\.com|\.co\.kr|\.net)",
+        r"공식|official|login|로그인|홈페이지|사이트|website",
+        r"고객센터|contact|support|customer\s?service",
+    ],
+    "informational": [
+        r"방법|how\s?to|what\s?is|why|when|where|who|which",
+        r"뜻|의미|정의|definition|meaning|guide|tutorial",
+        r"효과|부작용|증상|원인|차이|종류|type|cause|symptom|effect",
+        r"전후|before\s?and\s?after|결과|result",
+    ],
+}
+
+
+# ---------------------------------------------------------------------------
+# Dataclasses
+# ---------------------------------------------------------------------------
+
+
+@dataclass
+class OrganicKeyword:
+    """A keyword that a domain ranks for organically."""
+
+    keyword: str
+    position: int = 0
+    volume: int = 0
+    kd: float = 0.0
+    cpc: float = 0.0
+    url: str = ""
+    traffic: int = 0
+
+
+@dataclass
+class GapKeyword:
+    """A keyword gap between target and competitor(s)."""
+
+    keyword: str
+    volume: int = 0
+    kd: float = 0.0
+    cpc: float = 0.0
+    intent: str = "informational"
+    opportunity_score: float = 0.0
+    competitor_positions: dict[str, int] = field(default_factory=dict)
+    competitor_urls: dict[str, str] = field(default_factory=dict)
+    avg_competitor_position: float = 0.0
+
+    def to_dict(self) -> dict:
+        return asdict(self)
+
+
+@dataclass
+class GapAnalysisResult:
+    """Complete gap analysis result."""
+
+    target: str
+    competitors: list[str] = field(default_factory=list)
+    country: str = "kr"
+    total_gaps: int = 0
+    total_opportunity_volume: int = 0
+    gaps_by_intent: dict[str, int] = field(default_factory=dict)
+    top_opportunities: list[GapKeyword] = field(default_factory=list)
+    all_gaps: list[GapKeyword] = field(default_factory=list)
+    target_keyword_count: int = 0
+    competitor_keyword_counts: dict[str, int] = field(default_factory=dict)
+    timestamp: str = ""
+
+    def to_dict(self) -> dict:
+        return {
+            "target": self.target,
+            "competitors": self.competitors,
+            "country": self.country,
+            "total_gaps": self.total_gaps,
+            "total_opportunity_volume": self.total_opportunity_volume,
+            "gaps_by_intent": self.gaps_by_intent,
+            "top_opportunities": [g.to_dict() for g in self.top_opportunities],
+            "all_gaps": [g.to_dict() for g in self.all_gaps],
+            "target_keyword_count": self.target_keyword_count,
+            "competitor_keyword_counts": self.competitor_keyword_counts,
+            "timestamp": self.timestamp,
+        }
+
+
+# ---------------------------------------------------------------------------
+# MCP Helper
+# ---------------------------------------------------------------------------
+
+
+def call_mcp_tool(tool_name: str, params: dict) -> dict:
+    """
+    Call an Ahrefs MCP tool and return parsed JSON response.
+
+    In production this delegates to the MCP bridge. For standalone usage
+    it invokes the Claude CLI with the appropriate tool call.
+    """
+    logger.info(f"Calling MCP tool: {tool_name} with params: {json.dumps(params, ensure_ascii=False)}")
+
+    try:
+        cmd = [
+            "claude",
+            "--print",
+            "--output-format", "json",
+            "-p",
+            (
+                f"Call the tool mcp__claude_ai_Ahrefs__{tool_name} with these parameters: "
+                f"{json.dumps(params, ensure_ascii=False)}. Return ONLY the raw JSON result."
+            ),
+        ]
+        result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
+
+        if result.returncode != 0:
+            logger.warning(f"MCP tool {tool_name} returned non-zero exit code: {result.returncode}")
+            logger.debug(f"stderr: {result.stderr}")
+            return {"error": result.stderr, "keywords": [], "items": []}
+
+        try:
+            return json.loads(result.stdout)
+        except json.JSONDecodeError:
+            return {"raw": result.stdout, "keywords": [], "items": []}
+
+    except subprocess.TimeoutExpired:
+        logger.error(f"MCP tool {tool_name} timed out")
+        return {"error": "timeout", "keywords": [], "items": []}
+    except FileNotFoundError:
+        logger.warning("Claude CLI not found - returning empty result for standalone testing")
+        return {"keywords": [], "items": []}
+
+
+# ---------------------------------------------------------------------------
+# Utility functions
+# ---------------------------------------------------------------------------
+
+
+def extract_domain(url: str) -> str:
+    """Extract clean domain from URL."""
+    if not url.startswith(("http://", "https://")):
+        url = f"https://{url}"
+    parsed = urlparse(url)
+    domain = parsed.netloc or parsed.path
+    domain = domain.lower().strip("/")
+    if domain.startswith("www."):
+        domain = domain[4:]
+    return domain
+
+
+def classify_intent(keyword: str) -> str:
+    """Classify search intent based on keyword patterns."""
+    keyword_lower = keyword.lower().strip()
+    for intent, patterns in INTENT_PATTERNS.items():
+        for pattern in patterns:
+            if re.search(pattern, keyword_lower, re.IGNORECASE):
+                return intent
+    return "informational"
+
+
+# ---------------------------------------------------------------------------
+# KeywordGapAnalyzer
+# ---------------------------------------------------------------------------
+
+
+class KeywordGapAnalyzer:
+    """Analyze keyword gaps between a target site and its competitors."""
+
+    def __init__(self, country: str = "kr", min_volume: int = 0):
+        self.country = country
+        self.min_volume = min_volume
+
+    def get_organic_keywords(self, domain: str, limit: int = 1000) -> list[OrganicKeyword]:
+        """
+        Fetch organic keywords for a domain via Ahrefs site-explorer-organic-keywords.
+        Returns a list of OrganicKeyword entries.
+        """
+        clean_domain = extract_domain(domain)
+        logger.info(f"Fetching organic keywords for: {clean_domain} (limit={limit})")
+
+        result = call_mcp_tool("site-explorer-organic-keywords", {
+            "target": clean_domain,
+            "country": self.country,
+            "limit": limit,
+            "mode": "domain",
+        })
+
+        keywords: list[OrganicKeyword] = []
+        for item in result.get("keywords", result.get("items", [])):
+            if not isinstance(item, dict):
+                continue
+            kw = OrganicKeyword(
+                keyword=item.get("keyword", item.get("term", "")),
+                position=int(item.get("position", item.get("rank", 0)) or 0),
+                volume=int(item.get("volume", item.get("search_volume", 0)) or 0),
+                kd=float(item.get("keyword_difficulty", item.get("kd", 0)) or 0),
+                cpc=float(item.get("cpc", item.get("cost_per_click", 0)) or 0),
+                url=item.get("url", item.get("best_position_url", "")),
+                traffic=int(item.get("traffic", item.get("estimated_traffic", 0)) or 0),
+            )
+            if kw.keyword:
+                keywords.append(kw)
+
+        logger.info(f"Found {len(keywords)} organic keywords for {clean_domain}")
+        return keywords
+
+    def find_gaps(
+        self,
+        target_keywords: list[OrganicKeyword],
+        competitor_keyword_sets: dict[str, list[OrganicKeyword]],
+    ) -> list[GapKeyword]:
+        """
+        Identify keywords that competitors rank for but the target doesn't.
+
+        A gap keyword is one that appears in at least one competitor's keyword
+        set but not in the target's keyword set.
+        """
+        # Build target keyword set for fast lookup
+        target_kw_set: set[str] = {kw.keyword.lower().strip() for kw in target_keywords}
+
+        # Collect all competitor keywords with their positions
+        gap_map: dict[str, GapKeyword] = {}
+
+        for comp_domain, comp_keywords in competitor_keyword_sets.items():
+            for ckw in comp_keywords:
+                kw_lower = ckw.keyword.lower().strip()
+
+                # Skip if target already ranks for this keyword
+                if kw_lower in target_kw_set:
+                    continue
+
+                # Skip below minimum volume
+                if ckw.volume < self.min_volume:
+                    continue
+
+                if kw_lower not in gap_map:
+                    gap_map[kw_lower] = GapKeyword(
+                        keyword=ckw.keyword,
+                        volume=ckw.volume,
+                        kd=ckw.kd,
+                        cpc=ckw.cpc,
+                        intent=classify_intent(ckw.keyword),
+                        competitor_positions={},
+                        competitor_urls={},
+                    )
+
+                gap_map[kw_lower].competitor_positions[comp_domain] = ckw.position
+                gap_map[kw_lower].competitor_urls[comp_domain] = ckw.url
+
+                # Update volume/kd if higher from another competitor
+                if ckw.volume > gap_map[kw_lower].volume:
+                    gap_map[kw_lower].volume = ckw.volume
+                if ckw.kd > 0 and (gap_map[kw_lower].kd == 0 or ckw.kd < gap_map[kw_lower].kd):
+                    gap_map[kw_lower].kd = ckw.kd
+
+        gaps = list(gap_map.values())
+
+        # Calculate average competitor position for each gap
+        for gap in gaps:
+            positions = list(gap.competitor_positions.values())
+            gap.avg_competitor_position = round(
+                sum(positions) / len(positions), 1
+            ) if positions else 0.0
+
+        logger.info(f"Found {len(gaps)} keyword gaps")
+        return gaps
+
+    def score_opportunities(self, gaps: list[GapKeyword]) -> list[GapKeyword]:
+        """
+        Score each gap keyword by opportunity potential.
+
+        Formula:
+            opportunity_score = (volume_score * 0.4) + (kd_score * 0.3) +
+                                (position_score * 0.2) + (intent_score * 0.1)
+
+        Where:
+            - volume_score: normalized 0-100 based on max volume in set
+            - kd_score: inverted (lower KD = higher score), normalized 0-100
+            - position_score: based on avg competitor position (lower = easier to compete)
+            - intent_score: commercial/transactional get higher scores
+        """
+        if not gaps:
+            return gaps
+
+        # Find max volume for normalization
+        max_volume = max(g.volume for g in gaps) if gaps else 1
+        max_volume = max(max_volume, 1)
+
+        intent_scores = {
+            "transactional": 100,
+            "commercial": 80,
+            "informational": 40,
+            "navigational": 20,
+        }
+
+        for gap in gaps:
+            # Volume score (0-100)
+            volume_score = (gap.volume / max_volume) * 100
+
+            # KD score (inverted: low KD = high score)
+            kd_score = max(0, 100 - gap.kd)
+
+            # Position score (competitors ranking 1-10 means realistic opportunity)
+            if gap.avg_competitor_position <= 10:
+                position_score = 90
+            elif gap.avg_competitor_position <= 20:
+                position_score = 70
+            elif gap.avg_competitor_position <= 50:
+                position_score = 50
+            else:
+                position_score = 30
+
+            # Intent score
+            intent_score = intent_scores.get(gap.intent, 40)
+
+            # Combined score
+            gap.opportunity_score = round(
+                (volume_score * 0.4) +
+                (kd_score * 0.3) +
+                (position_score * 0.2) +
+                (intent_score * 0.1),
+                1,
+            )
+
+        # Sort by opportunity score descending
+        gaps.sort(key=lambda g: g.opportunity_score, reverse=True)
+
+        logger.info(f"Scored {len(gaps)} gap keywords by opportunity")
+        return gaps
+
+    def analyze(self, target_url: str, competitor_urls: list[str]) -> GapAnalysisResult:
+        """
+        Orchestrate full keyword gap analysis:
+        1. Fetch organic keywords for target
+        2. Fetch organic keywords for each competitor
+        3. Identify gaps
+        4. Score opportunities
+        5. Compile results
+        """
+        target_domain = extract_domain(target_url)
+        competitor_domains = [extract_domain(url) for url in competitor_urls]
+
+        logger.info(
+            f"Starting gap analysis: {target_domain} vs {', '.join(competitor_domains)}"
+        )
+
+        # Step 1: Fetch target keywords
+        target_keywords = self.get_organic_keywords(target_domain)
+
+        # Step 2: Fetch competitor keywords
+        competitor_keyword_sets: dict[str, list[OrganicKeyword]] = {}
+        competitor_keyword_counts: dict[str, int] = {}
+
+        for comp_domain in competitor_domains:
+            comp_keywords = self.get_organic_keywords(comp_domain)
+            competitor_keyword_sets[comp_domain] = comp_keywords
+            competitor_keyword_counts[comp_domain] = len(comp_keywords)
+
+        # Step 3: Find gaps
+        gaps = self.find_gaps(target_keywords, competitor_keyword_sets)
+
+        # Step 4: Score opportunities
+        scored_gaps = self.score_opportunities(gaps)
+
+        # Step 5: Calculate intent distribution
+        gaps_by_intent: dict[str, int] = {}
+        for gap in scored_gaps:
+            gaps_by_intent[gap.intent] = gaps_by_intent.get(gap.intent, 0) + 1
+
+        # Step 6: Compile result
+        result = GapAnalysisResult(
+            target=target_domain,
+            competitors=competitor_domains,
+            country=self.country,
+            total_gaps=len(scored_gaps),
+            total_opportunity_volume=sum(g.volume for g in scored_gaps),
+            gaps_by_intent=gaps_by_intent,
+            top_opportunities=scored_gaps[:50],
+            all_gaps=scored_gaps,
+            target_keyword_count=len(target_keywords),
+            competitor_keyword_counts=competitor_keyword_counts,
+            timestamp=datetime.now().isoformat(),
+        )
+
+        logger.info(
+            f"Gap analysis complete: {result.total_gaps} gaps found, "
+            f"total opportunity volume {result.total_opportunity_volume:,}"
+        )
+        return result
+
+
+# ---------------------------------------------------------------------------
+# Plain-text report formatter
+# ---------------------------------------------------------------------------
+
+
+def format_text_report(result: GapAnalysisResult) -> str:
+    """Format gap analysis result as a human-readable text report."""
+    lines: list[str] = []
+    lines.append("=" * 75)
+    lines.append(f"Keyword Gap Analysis Report")
+    lines.append(f"Target: {result.target}")
+    lines.append(f"Competitors: {', '.join(result.competitors)}")
+    lines.append(f"Country: {result.country.upper()} | Date: {result.timestamp[:10]}")
+    lines.append("=" * 75)
+    lines.append("")
+
+    # Overview
+    lines.append("## Overview")
+    lines.append(f"  Target keywords: {result.target_keyword_count:,}")
+    for comp, count in result.competitor_keyword_counts.items():
+        lines.append(f"  {comp} keywords: {count:,}")
+    lines.append(f"  Keyword gaps found: {result.total_gaps:,}")
+    lines.append(f"  Total opportunity volume: {result.total_opportunity_volume:,}")
+    lines.append("")
+
+    # Intent distribution
+    if result.gaps_by_intent:
+        lines.append("## Gaps by Intent")
+        for intent, count in sorted(result.gaps_by_intent.items(), key=lambda x: x[1], reverse=True):
+            pct = (count / result.total_gaps) * 100 if result.total_gaps else 0
+            lines.append(f"  {intent:<15}: {count:>5} ({pct:.1f}%)")
+        lines.append("")
+
+    # Top opportunities
+    if result.top_opportunities:
+        lines.append("## Top Opportunities (by score)")
+        header = f"  {'Keyword':<35} {'Vol':>8} {'KD':>6} {'Score':>7} {'Intent':<15} {'Competitors'}"
+        lines.append(header)
+        lines.append("  " + "-" * 90)
+
+        for gap in result.top_opportunities[:30]:
+            kw_display = gap.keyword[:33] if len(gap.keyword) > 33 else gap.keyword
+            comp_positions = ", ".join(
+                f"{d}:#{p}" for d, p in gap.competitor_positions.items()
+            )
+            comp_display = comp_positions[:30] if len(comp_positions) > 30 else comp_positions
+
+            lines.append(
+                f"  {kw_display:<35} {gap.volume:>8,} {gap.kd:>6.1f} "
+                f"{gap.opportunity_score:>7.1f} {gap.intent:<15} {comp_display}"
+            )
+        lines.append("")
+
+    # Quick wins (low KD, high volume)
+    quick_wins = [g for g in result.all_gaps if g.kd <= 30 and g.volume >= 100]
+    quick_wins.sort(key=lambda g: g.volume, reverse=True)
+    if quick_wins:
+        lines.append("## Quick Wins (KD <= 30, Volume >= 100)")
+        lines.append(f"  {'Keyword':<35} {'Vol':>8} {'KD':>6} {'Intent':<15}")
+        lines.append("  " + "-" * 64)
+        for gap in quick_wins[:20]:
+            kw_display = gap.keyword[:33] if len(gap.keyword) > 33 else gap.keyword
+            lines.append(
+                f"  {kw_display:<35} {gap.volume:>8,} {gap.kd:>6.1f} {gap.intent:<15}"
+            )
+        lines.append("")
+
+    return "\n".join(lines)
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Keyword Gap Analyzer - Find competitor keyword opportunities",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python keyword_gap_analyzer.py --target https://example.com --competitor https://comp.com --json
+  python keyword_gap_analyzer.py --target example.com --competitor comp1.com --competitor comp2.com --min-volume 100 --json
+  python keyword_gap_analyzer.py --target example.com --competitor comp.com --country us --output gaps.json
+        """,
+    )
+    parser.add_argument(
+        "--target",
+        required=True,
+        help="Target website URL or domain",
+    )
+    parser.add_argument(
+        "--competitor",
+        action="append",
+        required=True,
+        dest="competitors",
+        help="Competitor URL or domain (can be repeated)",
+    )
+    parser.add_argument(
+        "--country",
+        default="kr",
+        help="Target country code (default: kr)",
+    )
+    parser.add_argument(
+        "--min-volume",
+        type=int,
+        default=0,
+        help="Minimum search volume filter (default: 0)",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        dest="output_json",
+        help="Output results as JSON",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        default=None,
+        help="Write output to file (path)",
+    )
+    parser.add_argument(
+        "--verbose",
+        action="store_true",
+        help="Enable verbose/debug logging",
+    )
+
+    args = parser.parse_args()
+
+    if args.verbose:
+        logging.getLogger().setLevel(logging.DEBUG)
+
+    # Run analysis
+    analyzer = KeywordGapAnalyzer(
+        country=args.country,
+        min_volume=args.min_volume,
+    )
+    result = analyzer.analyze(args.target, args.competitors)
+
+    # Format output
+    if args.output_json:
+        output = json.dumps(result.to_dict(), ensure_ascii=False, indent=2)
+    else:
+        output = format_text_report(result)
+
+    # Write or print
+    if args.output:
+        with open(args.output, "w", encoding="utf-8") as f:
+            f.write(output)
+        logger.info(f"Output written to: {args.output}")
+    else:
+        print(output)
+
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())
diff --git a/custom-skills/19-seo-keyword-strategy/code/scripts/keyword_researcher.py b/custom-skills/19-seo-keyword-strategy/code/scripts/keyword_researcher.py
new file mode 100644
index 0000000..3cb3565
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/code/scripts/keyword_researcher.py
@@ -0,0 +1,656 @@
+"""
+Keyword Researcher - Seed keyword expansion, intent classification, and topic clustering
+========================================================================================
+Purpose: Expand seed keywords via Ahrefs APIs, classify search intent,
+         cluster topics, and support Korean market keyword discovery.
+Python: 3.10+
+"""
+
+import argparse
+import json
+import logging
+import re
+import subprocess
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from typing import Optional
+
+# ---------------------------------------------------------------------------
+# Logging
+# ---------------------------------------------------------------------------
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+logger = logging.getLogger("keyword_researcher")
+
+# ---------------------------------------------------------------------------
+# Constants - Korean suffix expansion
+# ---------------------------------------------------------------------------
+KOREAN_SUFFIXES: list[str] = [
+    "추천",
+    "가격",
+    "후기",
+    "잘하는곳",
+    "부작용",
+    "전후",
+    "비용",
+    "추천 병원",
+    "후기 블로그",
+    "방법",
+    "종류",
+    "비교",
+    "효과",
+    "주의사항",
+    "장단점",
+]
+
+# ---------------------------------------------------------------------------
+# Intent classification patterns
+# ---------------------------------------------------------------------------
+INTENT_PATTERNS: dict[str, list[str]] = {
+    "transactional": [
+        r"구매|구입|주문|buy|order|purchase|shop|deal|discount|coupon|할인|쿠폰",
+        r"예약|booking|reserve|sign\s?up|register|등록|신청",
+    ],
+    "commercial": [
+        r"가격|비용|얼마|price|cost|pricing|fee|요금",
+        r"추천|best|top\s?\d|review|비교|compare|vs|versus|후기|리뷰|평점|평가",
+        r"잘하는곳|잘하는|맛집|업체|병원|추천\s?병원",
+    ],
+    "navigational": [
+        r"^(www\.|http|\.com|\.co\.kr|\.net)",
+        r"공식|official|login|로그인|홈페이지|사이트|website",
+        r"고객센터|contact|support|customer\s?service",
+    ],
+    "informational": [
+        r"방법|how\s?to|what\s?is|why|when|where|who|which",
+        r"뜻|의미|정의|definition|meaning|guide|tutorial",
+        r"효과|부작용|증상|원인|차이|종류|type|cause|symptom|effect",
+        r"전후|before\s?and\s?after|결과|result",
+    ],
+}
+
+# ---------------------------------------------------------------------------
+# Dataclasses
+# ---------------------------------------------------------------------------
+
+
+@dataclass
+class KeywordEntry:
+    """Single keyword with its metrics and classification."""
+
+    keyword: str
+    volume: int = 0
+    kd: float = 0.0
+    cpc: float = 0.0
+    intent: str = "informational"
+    cluster: str = ""
+    source: str = ""
+    country_volumes: dict[str, int] = field(default_factory=dict)
+
+    def to_dict(self) -> dict:
+        data = asdict(self)
+        if not data["country_volumes"]:
+            del data["country_volumes"]
+        return data
+
+
+@dataclass
+class KeywordCluster:
+    """Group of semantically related keywords."""
+
+    topic: str
+    keywords: list[str] = field(default_factory=list)
+    total_volume: int = 0
+    avg_kd: float = 0.0
+    primary_intent: str = "informational"
+
+    def to_dict(self) -> dict:
+        return asdict(self)
+
+
+@dataclass
+class ResearchResult:
+    """Full research result container."""
+
+    seed_keyword: str
+    country: str
+    total_keywords: int = 0
+    total_volume: int = 0
+    clusters: list[KeywordCluster] = field(default_factory=list)
+    keywords: list[KeywordEntry] = field(default_factory=list)
+    timestamp: str = ""
+
+    def to_dict(self) -> dict:
+        return {
+            "seed_keyword": self.seed_keyword,
+            "country": self.country,
+            "total_keywords": self.total_keywords,
+            "total_volume": self.total_volume,
+            "clusters": [c.to_dict() for c in self.clusters],
+            "keywords": [k.to_dict() for k in self.keywords],
+            "timestamp": self.timestamp,
+        }
+
+
+# ---------------------------------------------------------------------------
+# MCP Helper - calls Ahrefs MCP tools via subprocess
+# ---------------------------------------------------------------------------
+
+
+def call_mcp_tool(tool_name: str, params: dict) -> dict:
+    """
+    Call an Ahrefs MCP tool and return parsed JSON response.
+
+    In production this delegates to the MCP bridge. For standalone usage
+    it invokes the Claude CLI with the appropriate tool call.
+    """
+    logger.info(f"Calling MCP tool: {tool_name} with params: {json.dumps(params, ensure_ascii=False)}")
+
+    try:
+        cmd = [
+            "claude",
+            "--print",
+            "--output-format", "json",
+            "-p",
+            f"Call the tool mcp__claude_ai_Ahrefs__{tool_name} with these parameters: {json.dumps(params, ensure_ascii=False)}. Return ONLY the raw JSON result.",
+        ]
+        result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
+
+        if result.returncode != 0:
+            logger.warning(f"MCP tool {tool_name} returned non-zero exit code: {result.returncode}")
+            logger.debug(f"stderr: {result.stderr}")
+            return {"error": result.stderr, "keywords": [], "items": []}
+
+        try:
+            return json.loads(result.stdout)
+        except json.JSONDecodeError:
+            return {"raw": result.stdout, "keywords": [], "items": []}
+
+    except subprocess.TimeoutExpired:
+        logger.error(f"MCP tool {tool_name} timed out")
+        return {"error": "timeout", "keywords": [], "items": []}
+    except FileNotFoundError:
+        logger.warning("Claude CLI not found - returning empty result for standalone testing")
+        return {"keywords": [], "items": []}
+
+
+# ---------------------------------------------------------------------------
+# KeywordResearcher
+# ---------------------------------------------------------------------------
+
+
+class KeywordResearcher:
+    """Expand seed keywords, classify intent, and cluster topics."""
+
+    def __init__(self, country: str = "kr", korean_suffixes: bool = False, compare_global: bool = False):
+        self.country = country
+        self.korean_suffixes = korean_suffixes
+        self.compare_global = compare_global
+        self._seen: set[str] = set()
+
+    # ---- Keyword expansion via Ahrefs MCP ----
+
+    def expand_keywords(self, seed: str) -> list[KeywordEntry]:
+        """
+        Expand a seed keyword using Ahrefs matching-terms, related-terms,
+        and search-suggestions endpoints.
+        """
+        all_keywords: list[KeywordEntry] = []
+
+        # 1. Matching terms
+        logger.info(f"Fetching matching terms for: {seed}")
+        matching = call_mcp_tool("keywords-explorer-matching-terms", {
+            "keyword": seed,
+            "country": self.country,
+            "limit": 100,
+        })
+        for item in matching.get("keywords", matching.get("items", [])):
+            kw = self._parse_keyword_item(item, source="matching-terms")
+            if kw and kw.keyword not in self._seen:
+                self._seen.add(kw.keyword)
+                all_keywords.append(kw)
+
+        # 2. Related terms
+        logger.info(f"Fetching related terms for: {seed}")
+        related = call_mcp_tool("keywords-explorer-related-terms", {
+            "keyword": seed,
+            "country": self.country,
+            "limit": 100,
+        })
+        for item in related.get("keywords", related.get("items", [])):
+            kw = self._parse_keyword_item(item, source="related-terms")
+            if kw and kw.keyword not in self._seen:
+                self._seen.add(kw.keyword)
+                all_keywords.append(kw)
+
+        # 3. Search suggestions
+        logger.info(f"Fetching search suggestions for: {seed}")
+        suggestions = call_mcp_tool("keywords-explorer-search-suggestions", {
+            "keyword": seed,
+            "country": self.country,
+            "limit": 50,
+        })
+        for item in suggestions.get("keywords", suggestions.get("items", [])):
+            kw = self._parse_keyword_item(item, source="search-suggestions")
+            if kw and kw.keyword not in self._seen:
+                self._seen.add(kw.keyword)
+                all_keywords.append(kw)
+
+        # 4. Add the seed itself if not already present
+        if seed not in self._seen:
+            self._seen.add(seed)
+            overview = call_mcp_tool("keywords-explorer-overview", {
+                "keyword": seed,
+                "country": self.country,
+            })
+            seed_entry = self._parse_keyword_item(overview, source="seed")
+            if seed_entry:
+                seed_entry.keyword = seed
+                all_keywords.insert(0, seed_entry)
+
+        logger.info(f"Expanded to {len(all_keywords)} keywords from Ahrefs APIs")
+        return all_keywords
+
+    def expand_korean_suffixes(self, seed: str) -> list[KeywordEntry]:
+        """
+        Generate keyword variations by appending common Korean suffixes.
+        Each variation is checked against Ahrefs for volume data.
+        """
+        suffix_keywords: list[KeywordEntry] = []
+
+        for suffix in KOREAN_SUFFIXES:
+            variation = f"{seed} {suffix}"
+            if variation in self._seen:
+                continue
+
+            logger.info(f"Checking Korean suffix variation: {variation}")
+            overview = call_mcp_tool("keywords-explorer-overview", {
+                "keyword": variation,
+                "country": self.country,
+            })
+            kw = self._parse_keyword_item(overview, source="korean-suffix")
+            if kw:
+                kw.keyword = variation
+                if kw.volume > 0:
+                    self._seen.add(variation)
+                    suffix_keywords.append(kw)
+            else:
+                # Even if no data, include as zero-volume for completeness
+                entry = KeywordEntry(
+                    keyword=variation,
+                    volume=0,
+                    kd=0.0,
+                    cpc=0.0,
+                    intent=self.classify_intent(variation),
+                    source="korean-suffix",
+                )
+                self._seen.add(variation)
+                suffix_keywords.append(entry)
+
+        logger.info(f"Korean suffix expansion yielded {len(suffix_keywords)} variations")
+        return suffix_keywords
+
+    def get_volume_by_country(self, keyword: str) -> dict[str, int]:
+        """
+        Get search volume breakdown by country for a keyword.
+        Useful for comparing Korean vs global demand.
+        """
+        logger.info(f"Fetching volume-by-country for: {keyword}")
+        result = call_mcp_tool("keywords-explorer-volume-by-country", {
+            "keyword": keyword,
+        })
+
+        volumes: dict[str, int] = {}
+        for item in result.get("countries", result.get("items", [])):
+            if isinstance(item, dict):
+                country_code = item.get("country", item.get("code", ""))
+                volume = item.get("volume", item.get("search_volume", 0))
+                if country_code and volume:
+                    volumes[country_code.lower()] = int(volume)
+
+        return volumes
+
+    # ---- Intent classification ----
+
+    def classify_intent(self, keyword: str) -> str:
+        """
+        Classify search intent based on keyword patterns.
+        Priority: transactional > commercial > navigational > informational
+        """
+        keyword_lower = keyword.lower().strip()
+
+        for intent, patterns in INTENT_PATTERNS.items():
+            for pattern in patterns:
+                if re.search(pattern, keyword_lower, re.IGNORECASE):
+                    return intent
+
+        return "informational"
+
+    # ---- Keyword clustering ----
+
+    def cluster_keywords(self, keywords: list[KeywordEntry]) -> list[KeywordCluster]:
+        """
+        Group keywords into topic clusters using shared n-gram tokens.
+        Uses a simple token overlap approach: keywords sharing significant
+        tokens (2+ character words) are grouped together.
+        """
+        if not keywords:
+            return []
+
+        # Extract meaningful tokens from each keyword
+        def tokenize(text: str) -> set[str]:
+            tokens = set()
+            for word in re.split(r"\s+", text.strip().lower()):
+                if len(word) >= 2:
+                    tokens.add(word)
+            return tokens
+
+        # Build token-to-keyword mapping
+        token_map: dict[str, list[int]] = {}
+        kw_tokens: list[set[str]] = []
+
+        for i, kw in enumerate(keywords):
+            tokens = tokenize(kw.keyword)
+            kw_tokens.append(tokens)
+            for token in tokens:
+                if token not in token_map:
+                    token_map[token] = []
+                token_map[token].append(i)
+
+        # Find the most common significant tokens (cluster anchors)
+        token_freq = sorted(token_map.items(), key=lambda x: len(x[1]), reverse=True)
+
+        assigned: set[int] = set()
+        clusters: list[KeywordCluster] = []
+
+        for token, indices in token_freq:
+            # Skip single-occurrence tokens or very common stop-like tokens
+            if len(indices) < 2:
+                continue
+
+            # Gather unassigned keywords that share this token
+            cluster_indices = [i for i in indices if i not in assigned]
+            if len(cluster_indices) < 2:
+                continue
+
+            # Create the cluster
+            cluster_kws = [keywords[i].keyword for i in cluster_indices]
+            cluster_volumes = [keywords[i].volume for i in cluster_indices]
+            cluster_kds = [keywords[i].kd for i in cluster_indices]
+            cluster_intents = [keywords[i].intent for i in cluster_indices]
+
+            # Determine primary intent by frequency
+            intent_counts: dict[str, int] = {}
+            for intent in cluster_intents:
+                intent_counts[intent] = intent_counts.get(intent, 0) + 1
+            primary_intent = max(intent_counts, key=intent_counts.get)
+
+            cluster = KeywordCluster(
+                topic=token,
+                keywords=cluster_kws,
+                total_volume=sum(cluster_volumes),
+                avg_kd=round(sum(cluster_kds) / len(cluster_kds), 1) if cluster_kds else 0.0,
+                primary_intent=primary_intent,
+            )
+            clusters.append(cluster)
+
+            for i in cluster_indices:
+                assigned.add(i)
+                keywords[i].cluster = token
+
+        # Assign unclustered keywords to an "other" cluster
+        unclustered = [i for i in range(len(keywords)) if i not in assigned]
+        if unclustered:
+            other_kws = [keywords[i].keyword for i in unclustered]
+            other_volumes = [keywords[i].volume for i in unclustered]
+            other_kds = [keywords[i].kd for i in unclustered]
+
+            other_cluster = KeywordCluster(
+                topic="(unclustered)",
+                keywords=other_kws,
+                total_volume=sum(other_volumes),
+                avg_kd=round(sum(other_kds) / len(other_kds), 1) if other_kds else 0.0,
+                primary_intent="informational",
+            )
+            clusters.append(other_cluster)
+
+            for i in unclustered:
+                keywords[i].cluster = "(unclustered)"
+
+        # Sort clusters by total volume descending
+        clusters.sort(key=lambda c: c.total_volume, reverse=True)
+
+        logger.info(f"Clustered {len(keywords)} keywords into {len(clusters)} clusters")
+        return clusters
+
+    # ---- Full analysis orchestration ----
+
+    def analyze(self, seed_keyword: str) -> ResearchResult:
+        """
+        Orchestrate a full keyword research analysis:
+        1. Expand seed via Ahrefs
+        2. Optionally expand Korean suffixes
+        3. Classify intent for all keywords
+        4. Optionally fetch volume-by-country
+        5. Cluster keywords into topics
+        6. Compile results
+        """
+        logger.info(f"Starting keyword research for: {seed_keyword} (country={self.country})")
+
+        # Step 1: Expand keywords
+        keywords = self.expand_keywords(seed_keyword)
+
+        # Step 2: Korean suffix expansion
+        if self.korean_suffixes:
+            suffix_keywords = self.expand_korean_suffixes(seed_keyword)
+            keywords.extend(suffix_keywords)
+
+        # Step 3: Classify intent for all keywords
+        for kw in keywords:
+            if not kw.intent or kw.intent == "informational":
+                kw.intent = self.classify_intent(kw.keyword)
+
+        # Step 4: Volume-by-country comparison
+        if self.compare_global and keywords:
+            # Fetch for the seed and top volume keywords
+            top_keywords = sorted(keywords, key=lambda k: k.volume, reverse=True)[:10]
+            for kw in top_keywords:
+                volumes = self.get_volume_by_country(kw.keyword)
+                kw.country_volumes = volumes
+
+        # Step 5: Cluster keywords
+        clusters = self.cluster_keywords(keywords)
+
+        # Step 6: Compile result
+        result = ResearchResult(
+            seed_keyword=seed_keyword,
+            country=self.country,
+            total_keywords=len(keywords),
+            total_volume=sum(kw.volume for kw in keywords),
+            clusters=clusters,
+            keywords=sorted(keywords, key=lambda k: k.volume, reverse=True),
+            timestamp=datetime.now().isoformat(),
+        )
+
+        logger.info(
+            f"Research complete: {result.total_keywords} keywords, "
+            f"{len(result.clusters)} clusters, "
+            f"total volume {result.total_volume}"
+        )
+        return result
+
+    # ---- Internal helpers ----
+
+    def _parse_keyword_item(self, item: dict, source: str = "") -> Optional[KeywordEntry]:
+        """Parse an Ahrefs API response item into a KeywordEntry."""
+        if not item or "error" in item:
+            return None
+
+        keyword = item.get("keyword", item.get("term", item.get("query", "")))
+        if not keyword:
+            return None
+
+        volume = int(item.get("volume", item.get("search_volume", 0)) or 0)
+        kd = float(item.get("keyword_difficulty", item.get("kd", 0)) or 0)
+        cpc = float(item.get("cpc", item.get("cost_per_click", 0)) or 0)
+
+        return KeywordEntry(
+            keyword=keyword,
+            volume=volume,
+            kd=round(kd, 1),
+            cpc=round(cpc, 2),
+            intent=self.classify_intent(keyword),
+            source=source,
+        )
+
+
+# ---------------------------------------------------------------------------
+# Plain-text report formatter
+# ---------------------------------------------------------------------------
+
+
+def format_text_report(result: ResearchResult) -> str:
+    """Format research result as a human-readable text report."""
+    lines: list[str] = []
+    lines.append("=" * 70)
+    lines.append(f"Keyword Strategy Report: {result.seed_keyword}")
+    lines.append(f"Country: {result.country.upper()} | Date: {result.timestamp[:10]}")
+    lines.append("=" * 70)
+    lines.append("")
+
+    lines.append("## Overview")
+    lines.append(f"  Total keywords discovered: {result.total_keywords}")
+    lines.append(f"  Topic clusters: {len(result.clusters)}")
+    lines.append(f"  Total search volume: {result.total_volume:,}")
+    lines.append("")
+
+    # Clusters summary
+    if result.clusters:
+        lines.append("## Top Clusters")
+        lines.append(f"  {'Cluster':<25} {'Keywords':>8} {'Volume':>10} {'Avg KD':>8} {'Intent':<15}")
+        lines.append("  " + "-" * 66)
+        for cluster in result.clusters[:15]:
+            lines.append(
+                f"  {cluster.topic:<25} {len(cluster.keywords):>8} "
+                f"{cluster.total_volume:>10,} {cluster.avg_kd:>8.1f} "
+                f"{cluster.primary_intent:<15}"
+            )
+        lines.append("")
+
+    # Top keywords
+    if result.keywords:
+        lines.append("## Top Keywords (by volume)")
+        lines.append(f"  {'Keyword':<40} {'Vol':>8} {'KD':>6} {'CPC':>7} {'Intent':<15} {'Cluster':<15}")
+        lines.append("  " + "-" * 91)
+        for kw in result.keywords[:30]:
+            kw_display = kw.keyword[:38] if len(kw.keyword) > 38 else kw.keyword
+            cluster_display = kw.cluster[:13] if len(kw.cluster) > 13 else kw.cluster
+            lines.append(
+                f"  {kw_display:<40} {kw.volume:>8,} {kw.kd:>6.1f} "
+                f"{kw.cpc:>7.2f} {kw.intent:<15} {cluster_display:<15}"
+            )
+        lines.append("")
+
+    # Intent distribution
+    intent_dist: dict[str, int] = {}
+    for kw in result.keywords:
+        intent_dist[kw.intent] = intent_dist.get(kw.intent, 0) + 1
+    if intent_dist:
+        lines.append("## Intent Distribution")
+        for intent, count in sorted(intent_dist.items(), key=lambda x: x[1], reverse=True):
+            pct = (count / len(result.keywords)) * 100 if result.keywords else 0
+            lines.append(f"  {intent:<15}: {count:>5} ({pct:.1f}%)")
+        lines.append("")
+
+    return "\n".join(lines)
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Keyword Researcher - Expand, classify, and cluster keywords",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python keyword_researcher.py --keyword "치과 임플란트" --country kr --json
+  python keyword_researcher.py --keyword "dental implant" --compare-global --json
+  python keyword_researcher.py --keyword "치과 임플란트" --korean-suffixes --output report.json
+        """,
+    )
+    parser.add_argument(
+        "--keyword",
+        required=True,
+        help="Seed keyword to expand and research",
+    )
+    parser.add_argument(
+        "--country",
+        default="kr",
+        help="Target country code (default: kr)",
+    )
+    parser.add_argument(
+        "--korean-suffixes",
+        action="store_true",
+        help="Enable Korean suffix expansion (추천, 가격, 후기, etc.)",
+    )
+    parser.add_argument(
+        "--compare-global",
+        action="store_true",
+        help="Fetch volume-by-country comparison for top keywords",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        dest="output_json",
+        help="Output results as JSON",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        default=None,
+        help="Write output to file (path)",
+    )
+    parser.add_argument(
+        "--verbose",
+        action="store_true",
+        help="Enable verbose/debug logging",
+    )
+
+    args = parser.parse_args()
+
+    if args.verbose:
+        logging.getLogger().setLevel(logging.DEBUG)
+
+    # Run analysis
+    researcher = KeywordResearcher(
+        country=args.country,
+        korean_suffixes=args.korean_suffixes,
+        compare_global=args.compare_global,
+    )
+    result = researcher.analyze(args.keyword)
+
+    # Format output
+    if args.output_json:
+        output = json.dumps(result.to_dict(), ensure_ascii=False, indent=2)
+    else:
+        output = format_text_report(result)
+
+    # Write or print
+    if args.output:
+        with open(args.output, "w", encoding="utf-8") as f:
+            f.write(output)
+        logger.info(f"Output written to: {args.output}")
+    else:
+        print(output)
+
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())
diff --git a/custom-skills/19-seo-keyword-strategy/code/scripts/requirements.txt b/custom-skills/19-seo-keyword-strategy/code/scripts/requirements.txt
new file mode 100644
index 0000000..ce98fae
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/code/scripts/requirements.txt
@@ -0,0 +1,20 @@
+# 19-seo-keyword-strategy dependencies
+# Install: pip install -r requirements.txt
+
+# HTTP & Async
+requests>=2.31.0
+aiohttp>=3.9.0
+
+# Data Processing
+pandas>=2.1.0
+
+# NLP / Text Similarity
+scikit-learn>=1.3.0
+
+# Async & Retry
+tenacity>=8.2.0
+tqdm>=4.66.0
+
+# Environment & CLI
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/19-seo-keyword-strategy/desktop/SKILL.md b/custom-skills/19-seo-keyword-strategy/desktop/SKILL.md
new file mode 100644
index 0000000..5db8269
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/desktop/SKILL.md
@@ -0,0 +1,91 @@
+---
+name: seo-keyword-strategy
+description: |
+  Keyword strategy and research for SEO campaigns. Triggers: keyword research, keyword analysis, keyword gap, search volume, keyword clustering, intent classification.
+---
+
+# SEO Keyword Strategy & Research
+
+## Purpose
+
+Expand seed keywords, classify search intent, cluster topics, and identify competitor keyword gaps for comprehensive keyword strategy development.
+
+## Core Capabilities
+
+1. **Keyword Expansion** - Matching terms, related terms, search suggestions
+2. **Korean Market** - Suffix expansion, Naver autocomplete, Korean intent patterns
+3. **Intent Classification** - Informational, navigational, commercial, transactional
+4. **Topic Clustering** - Group keywords into semantic clusters
+5. **Gap Analysis** - Find competitor keywords missing from target site
+
+## MCP Tool Usage
+
+### Ahrefs for Keyword Data
+```
+mcp__ahrefs__keywords-explorer-overview: Get keyword metrics
+mcp__ahrefs__keywords-explorer-matching-terms: Find keyword variations
+mcp__ahrefs__keywords-explorer-related-terms: Discover related keywords
+mcp__ahrefs__keywords-explorer-search-suggestions: Autocomplete suggestions
+mcp__ahrefs__keywords-explorer-volume-by-country: Country volume comparison
+mcp__ahrefs__site-explorer-organic-keywords: Competitor keyword rankings
+```
+
+### Web Search for Naver Discovery
+```
+WebSearch: Naver autocomplete and trend discovery
+```
+
+## Workflow
+
+### 1. Seed Keyword Expansion
+1. Input seed keyword (Korean or English)
+2. Query Ahrefs matching-terms and related-terms
+3. Get search suggestions for long-tail variations
+4. Apply Korean suffix expansion if Korean market
+5. Deduplicate and merge results
+
+### 2. Intent Classification & Clustering
+1. Classify each keyword by search intent
+2. Group keywords into topic clusters
+3. Identify pillar topics and supporting terms
+4. Calculate cluster-level metrics (total volume, avg KD)
+
+### 3. Gap Analysis
+1. Pull organic keywords for target and competitors
+2. Identify keywords present in competitors but missing from target
+3. Score opportunities by volume/difficulty ratio
+4. Prioritize by intent alignment with business goals
+
+## Output Format
+
+```markdown
+## Keyword Strategy Report: [seed keyword]
+
+### Overview
+- Total keywords discovered: [count]
+- Topic clusters: [count]
+- Total search volume: [sum]
+
+### Top Clusters
+| Cluster | Keywords | Total Volume | Avg KD |
+|---------|----------|-------------|--------|
+| ... | ... | ... | ... |
+
+### Top Opportunities
+| Keyword | Volume | KD | Intent | Cluster |
+|---------|--------|-----|--------|---------|
+| ... | ... | ... | ... | ... |
+
+### Keyword Gaps (vs competitors)
+| Keyword | Volume | Competitor Position | Opportunity Score |
+|---------|--------|-------------------|-------------------|
+| ... | ... | ... | ... |
+```
+
+## Notion Output (Required)
+
+All audit reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category, Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: KW-YYYYMMDD-NNN
diff --git a/custom-skills/19-seo-keyword-strategy/desktop/skill.yaml b/custom-skills/19-seo-keyword-strategy/desktop/skill.yaml
new file mode 100644
index 0000000..0bf39f0
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/desktop/skill.yaml
@@ -0,0 +1,9 @@
+name: seo-keyword-strategy
+description: |
+  Keyword strategy and research for SEO campaigns. Triggers: keyword research, keyword analysis, keyword gap, search volume, keyword clustering, intent classification.
+
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/19-seo-keyword-strategy/desktop/tools/ahrefs.md b/custom-skills/19-seo-keyword-strategy/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..5d918e0
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/desktop/tools/ahrefs.md
@@ -0,0 +1,15 @@
+# Ahrefs
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/19-seo-keyword-strategy/desktop/tools/notion.md b/custom-skills/19-seo-keyword-strategy/desktop/tools/notion.md
new file mode 100644
index 0000000..153e086
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/desktop/tools/notion.md
@@ -0,0 +1,15 @@
+# Notion
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/19-seo-keyword-strategy/desktop/tools/websearch.md b/custom-skills/19-seo-keyword-strategy/desktop/tools/websearch.md
new file mode 100644
index 0000000..36ef238
--- /dev/null
+++ b/custom-skills/19-seo-keyword-strategy/desktop/tools/websearch.md
@@ -0,0 +1,15 @@
+# WebSearch
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/20-seo-serp-analysis/code/CLAUDE.md b/custom-skills/20-seo-serp-analysis/code/CLAUDE.md
new file mode 100644
index 0000000..9bf0a95
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/code/CLAUDE.md
@@ -0,0 +1,132 @@
+# CLAUDE.md
+
+## Overview
+
+SERP analysis tool for understanding search result landscapes. Detects Google SERP features (featured snippets, PAA, knowledge panels, local pack, video, ads), analyzes Naver SERP composition (blog, cafe, knowledge iN, Smart Store, brand zone, VIEW tab), maps competitor positions, and scores SERP feature opportunities.
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Google SERP analysis
+python scripts/serp_analyzer.py --keyword "치과 임플란트" --country kr --json
+
+# Naver SERP analysis
+python scripts/naver_serp_analyzer.py --keyword "치과 임플란트" --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `serp_analyzer.py` | Google SERP feature detection and competitor mapping | SERP features, competitor positions, opportunity scores |
+| `naver_serp_analyzer.py` | Naver SERP composition analysis | Section distribution, content type mapping |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## SERP Analyzer (Google)
+
+```bash
+# Single keyword analysis
+python scripts/serp_analyzer.py --keyword "dental implant cost" --json
+
+# Korean market
+python scripts/serp_analyzer.py --keyword "치과 임플란트 가격" --country kr --json
+
+# Multiple keywords from file
+python scripts/serp_analyzer.py --keywords-file keywords.txt --country kr --json
+
+# Output to file
+python scripts/serp_analyzer.py --keyword "dental implant" --output serp_report.json
+```
+
+**Capabilities**:
+- SERP feature detection (featured snippet, PAA, knowledge panel, local pack, video carousel, ads, image pack, site links)
+- Competitor position mapping per keyword
+- Content type distribution analysis (blog, product, service, news, video)
+- SERP feature opportunity scoring
+- Search intent validation from SERP composition
+- SERP volatility assessment
+
+## Naver SERP Analyzer
+
+```bash
+# Analyze Naver search results
+python scripts/naver_serp_analyzer.py --keyword "치과 임플란트" --json
+
+# Analyze multiple keywords
+python scripts/naver_serp_analyzer.py --keywords-file keywords.txt --json
+```
+
+**Capabilities**:
+- Naver section detection (블로그, 카페, 지식iN, 스마트스토어, 브랜드존, VIEW탭)
+- Section priority mapping (which sections appear above fold)
+- Content type distribution per section
+- Brand zone presence detection
+- VIEW tab content analysis
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `serp-overview` | Get SERP results for a keyword |
+| `keywords-explorer-overview` | Get keyword metrics and SERP features |
+| `site-explorer-organic-keywords` | Map competitor positions |
+
+## Output Format
+
+```json
+{
+  "keyword": "치과 임플란트",
+  "country": "kr",
+  "serp_features": {
+    "featured_snippet": true,
+    "people_also_ask": true,
+    "local_pack": true,
+    "knowledge_panel": false,
+    "video_carousel": false,
+    "ads_top": 3,
+    "ads_bottom": 2
+  },
+  "competitors": [
+    {
+      "position": 1,
+      "url": "https://example.com/page",
+      "domain": "example.com",
+      "title": "...",
+      "content_type": "service_page"
+    }
+  ],
+  "opportunity_score": 72,
+  "intent_signals": "commercial",
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Audited website URL |
+| Category | Select | SERP Analysis |
+| Priority | Select | Based on opportunity score |
+| Found Date | Date | Analysis date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: SERP-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., SERP, Featured Snippet, PAA)
+- URLs and code remain unchanged
diff --git a/custom-skills/20-seo-serp-analysis/code/scripts/base_client.py b/custom-skills/20-seo-serp-analysis/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/20-seo-serp-analysis/code/scripts/naver_serp_analyzer.py b/custom-skills/20-seo-serp-analysis/code/scripts/naver_serp_analyzer.py
new file mode 100644
index 0000000..cbd1983
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/code/scripts/naver_serp_analyzer.py
@@ -0,0 +1,682 @@
+"""
+Naver SERP Analyzer - Naver search result composition analysis
+==============================================================
+Purpose: Analyze Naver SERP section distribution, content type mapping,
+         brand zone detection, and VIEW tab content analysis.
+Python: 3.10+
+
+Usage:
+    python naver_serp_analyzer.py --keyword "치과 임플란트" --json
+    python naver_serp_analyzer.py --keywords-file keywords.txt --json
+    python naver_serp_analyzer.py --keyword "치과 임플란트" --output naver_report.json
+"""
+
+import argparse
+import json
+import logging
+import re
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from pathlib import Path
+from typing import Any
+
+import requests
+from bs4 import BeautifulSoup
+from rich.console import Console
+from rich.table import Table
+
+# ---------------------------------------------------------------------------
+# Logging
+# ---------------------------------------------------------------------------
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+logger = logging.getLogger(__name__)
+console = Console()
+
+# ---------------------------------------------------------------------------
+# Constants - Naver SERP Section Identifiers
+# ---------------------------------------------------------------------------
+
+# CSS class / id patterns used to detect Naver SERP sections
+NAVER_SECTION_SELECTORS: dict[str, list[str]] = {
+    "blog": [
+        "sp_blog",
+        "blog_widget",
+        "sc_new.sp_blog",
+        "api_subject_blog",
+        "type_blog",
+        "blog_exact",
+    ],
+    "cafe": [
+        "sp_cafe",
+        "cafe_widget",
+        "sc_new.sp_cafe",
+        "api_subject_cafe",
+        "type_cafe",
+    ],
+    "knowledge_in": [
+        "sp_kin",
+        "kin_widget",
+        "sc_new.sp_kin",
+        "api_subject_kin",
+        "type_kin",
+        "nx_kin",
+    ],
+    "smart_store": [
+        "sp_nshop",
+        "shopping_widget",
+        "sc_new.sp_nshop",
+        "api_subject_shopping",
+        "type_shopping",
+        "smartstore",
+    ],
+    "brand_zone": [
+        "sp_brand",
+        "brand_area",
+        "brand_zone",
+        "type_brand",
+        "sc_new.sp_brand",
+    ],
+    "view_tab": [
+        "sp_view",
+        "view_widget",
+        "sc_new.sp_view",
+        "type_view",
+        "api_subject_view",
+    ],
+    "news": [
+        "sp_nnews",
+        "news_widget",
+        "sc_new.sp_nnews",
+        "api_subject_news",
+        "type_news",
+        "group_news",
+    ],
+    "encyclopedia": [
+        "sp_encyclopedia",
+        "sc_new.sp_encyclopedia",
+        "api_subject_encyclopedia",
+        "type_encyclopedia",
+        "nx_encyclopedia",
+    ],
+    "image": [
+        "sp_image",
+        "image_widget",
+        "sc_new.sp_image",
+        "api_subject_image",
+        "type_image",
+    ],
+    "video": [
+        "sp_video",
+        "video_widget",
+        "sc_new.sp_video",
+        "api_subject_video",
+        "type_video",
+    ],
+    "place": [
+        "sp_local",
+        "local_widget",
+        "sc_new.sp_local",
+        "type_place",
+        "place_section",
+        "loc_map",
+    ],
+    "ad": [
+        "sp_nad",
+        "sp_tad",
+        "ad_section",
+        "type_powerlink",
+        "type_ad",
+        "nx_ad",
+    ],
+}
+
+# Section display names in Korean
+SECTION_DISPLAY_NAMES: dict[str, str] = {
+    "blog": "블로그",
+    "cafe": "카페",
+    "knowledge_in": "지식iN",
+    "smart_store": "스마트스토어",
+    "brand_zone": "브랜드존",
+    "view_tab": "VIEW",
+    "news": "뉴스",
+    "encyclopedia": "백과사전",
+    "image": "이미지",
+    "video": "동영상",
+    "place": "플레이스",
+    "ad": "광고",
+}
+
+# Default headers for Naver requests
+NAVER_HEADERS = {
+    "User-Agent": (
+        "Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
+        "AppleWebKit/537.36 (KHTML, like Gecko) "
+        "Chrome/120.0.0.0 Safari/537.36"
+    ),
+    "Accept-Language": "ko-KR,ko;q=0.9,en-US;q=0.8,en;q=0.7",
+    "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
+}
+
+
+# ---------------------------------------------------------------------------
+# Data Classes
+# ---------------------------------------------------------------------------
+
+
+@dataclass
+class NaverSection:
+    """A detected section within Naver SERP."""
+
+    section_type: str  # blog, cafe, knowledge_in, smart_store, etc.
+    display_name: str = ""
+    position: int = 0  # Order of appearance (1-based)
+    item_count: int = 0  # Number of items in the section
+    is_above_fold: bool = False  # Appears within first ~3 sections
+    has_more_link: bool = False  # Section has "more results" link
+    raw_html_snippet: str = ""  # Short HTML snippet for debugging
+
+    def __post_init__(self):
+        if not self.display_name:
+            self.display_name = SECTION_DISPLAY_NAMES.get(
+                self.section_type, self.section_type
+            )
+
+
+@dataclass
+class NaverSerpResult:
+    """Complete Naver SERP analysis result for a keyword."""
+
+    keyword: str
+    sections: list[NaverSection] = field(default_factory=list)
+    section_order: list[str] = field(default_factory=list)
+    brand_zone_present: bool = False
+    brand_zone_brand: str = ""
+    total_sections: int = 0
+    above_fold_sections: list[str] = field(default_factory=list)
+    ad_count: int = 0
+    dominant_section: str = ""
+    has_view_tab: bool = False
+    has_place_section: bool = False
+    timestamp: str = ""
+
+    def __post_init__(self):
+        if not self.timestamp:
+            self.timestamp = datetime.now().isoformat()
+
+
+# ---------------------------------------------------------------------------
+# Naver SERP Analyzer
+# ---------------------------------------------------------------------------
+
+
+class NaverSerpAnalyzer:
+    """Analyzes Naver search result page composition."""
+
+    NAVER_SEARCH_URL = "https://search.naver.com/search.naver"
+
+    def __init__(self, timeout: int = 30):
+        self.timeout = timeout
+        self.logger = logging.getLogger(self.__class__.__name__)
+        self.session = requests.Session()
+        self.session.headers.update(NAVER_HEADERS)
+
+    # ----- Data Fetching -----
+
+    def fetch_serp(self, keyword: str) -> str:
+        """
+        Fetch Naver search results HTML for a given keyword.
+
+        Returns the raw HTML string of the search results page.
+        """
+        self.logger.info(f"Fetching Naver SERP for '{keyword}'")
+
+        params = {
+            "where": "nexearch",
+            "sm": "top_hty",
+            "fbm": "0",
+            "ie": "utf8",
+            "query": keyword,
+        }
+
+        try:
+            response = self.session.get(
+                self.NAVER_SEARCH_URL,
+                params=params,
+                timeout=self.timeout,
+            )
+            response.raise_for_status()
+            self.logger.info(
+                f"Fetched {len(response.text):,} bytes "
+                f"(status={response.status_code})"
+            )
+            return response.text
+
+        except requests.RequestException as exc:
+            self.logger.error(f"Failed to fetch Naver SERP: {exc}")
+            return ""
+
+    # ----- Section Detection -----
+
+    def detect_sections(self, html: str) -> list[NaverSection]:
+        """
+        Identify Naver SERP sections from HTML structure.
+
+        Scans the HTML for known CSS class names and IDs that correspond
+        to Naver's SERP section types.
+        """
+        if not html:
+            return []
+
+        soup = BeautifulSoup(html, "lxml")
+        sections: list[NaverSection] = []
+        position = 0
+
+        # Strategy 1: Look for section containers with known class names
+        # Naver uses <div class="sc_new sp_XXX"> and <section> elements
+        all_sections = soup.find_all(
+            ["div", "section"],
+            class_=re.compile(
+                r"(sc_new|api_subject|sp_|type_|_widget|group_|nx_)"
+            ),
+        )
+
+        seen_types: set[str] = set()
+
+        for element in all_sections:
+            classes = " ".join(element.get("class", []))
+            element_id = element.get("id", "")
+            search_text = f"{classes} {element_id}".lower()
+
+            for section_type, selectors in NAVER_SECTION_SELECTORS.items():
+                if section_type in seen_types:
+                    continue
+
+                matched = False
+                for selector in selectors:
+                    if selector.lower() in search_text:
+                        matched = True
+                        break
+
+                if matched:
+                    position += 1
+                    seen_types.add(section_type)
+
+                    # Count items within the section
+                    item_count = self._count_section_items(element, section_type)
+
+                    # Check for "more" link
+                    has_more = bool(
+                        element.find("a", class_=re.compile(r"(more|_more|btn_more)"))
+                        or element.find("a", string=re.compile(r"(더보기|전체보기)"))
+                    )
+
+                    # Get short HTML snippet for debugging
+                    snippet = str(element)[:200] if element else ""
+
+                    section = NaverSection(
+                        section_type=section_type,
+                        position=position,
+                        item_count=item_count,
+                        is_above_fold=(position <= 3),
+                        has_more_link=has_more,
+                        raw_html_snippet=snippet,
+                    )
+                    sections.append(section)
+
+        # Strategy 2: Fallback - scan entire HTML text for section markers
+        if not sections:
+            self.logger.warning(
+                "No sections found via DOM parsing; "
+                "falling back to text pattern matching"
+            )
+            sections = self._fallback_text_detection(html)
+
+        return sections
+
+    def _count_section_items(self, element: Any, section_type: str) -> int:
+        """Count the number of result items within a section element."""
+        # Common item container patterns
+        item_selectors = [
+            "li",
+            ".api_txt_lines",
+            ".total_tit",
+            ".detail_box",
+            ".item",
+            ".lst_total > li",
+        ]
+
+        for selector in item_selectors:
+            items = element.select(selector)
+            if items and len(items) > 0:
+                return len(items)
+
+        # Fallback: count links that look like results
+        links = element.find_all("a", href=True)
+        result_links = [
+            a
+            for a in links
+            if a.get("href", "").startswith("http")
+            and "naver.com/search" not in a.get("href", "")
+        ]
+        return len(result_links) if result_links else 0
+
+    def _fallback_text_detection(self, html: str) -> list[NaverSection]:
+        """Detect sections by scanning raw HTML text for known markers."""
+        sections: list[NaverSection] = []
+        position = 0
+        html_lower = html.lower()
+
+        for section_type, selectors in NAVER_SECTION_SELECTORS.items():
+            for selector in selectors:
+                if selector.lower() in html_lower:
+                    position += 1
+                    sections.append(
+                        NaverSection(
+                            section_type=section_type,
+                            position=position,
+                            item_count=0,
+                            is_above_fold=(position <= 3),
+                        )
+                    )
+                    break
+
+        return sections
+
+    # ----- Section Priority Analysis -----
+
+    def analyze_section_priority(
+        self, sections: list[NaverSection]
+    ) -> list[str]:
+        """
+        Determine above-fold section order.
+
+        Returns ordered list of section types that appear in the first
+        visible area of the SERP (approximately top 3 sections).
+        """
+        sorted_sections = sorted(sections, key=lambda s: s.position)
+        above_fold = [s.section_type for s in sorted_sections if s.is_above_fold]
+        return above_fold
+
+    # ----- Brand Zone Detection -----
+
+    def check_brand_zone(self, html: str) -> tuple[bool, str]:
+        """
+        Detect brand zone presence and extract brand name if available.
+
+        Returns (is_present, brand_name).
+        """
+        if not html:
+            return False, ""
+
+        soup = BeautifulSoup(html, "lxml")
+
+        # Look for brand zone container
+        brand_selectors = [
+            "sp_brand",
+            "brand_area",
+            "brand_zone",
+            "type_brand",
+        ]
+
+        for selector in brand_selectors:
+            brand_el = soup.find(
+                ["div", "section"],
+                class_=re.compile(selector, re.IGNORECASE),
+            )
+            if brand_el:
+                # Try to extract brand name from the section
+                brand_name = ""
+                title_el = brand_el.find(
+                    ["h2", "h3", "strong", "a"],
+                    class_=re.compile(r"(tit|title|name|brand)", re.IGNORECASE),
+                )
+                if title_el:
+                    brand_name = title_el.get_text(strip=True)
+
+                return True, brand_name
+
+        # Text-based fallback
+        if "brand_zone" in html.lower() or "sp_brand" in html.lower():
+            return True, ""
+
+        return False, ""
+
+    # ----- Dominant Section -----
+
+    def _find_dominant_section(self, sections: list[NaverSection]) -> str:
+        """Find the section with the most items (excluding ads)."""
+        non_ad = [s for s in sections if s.section_type != "ad"]
+        if not non_ad:
+            return ""
+        return max(non_ad, key=lambda s: s.item_count).section_type
+
+    # ----- Main Analysis Orchestrator -----
+
+    def analyze(self, keyword: str) -> NaverSerpResult:
+        """
+        Orchestrate full Naver SERP analysis for a single keyword.
+
+        Steps:
+        1. Fetch Naver search results page
+        2. Detect SERP sections
+        3. Analyze section priority
+        4. Check brand zone presence
+        5. Compile results
+        """
+        html = self.fetch_serp(keyword)
+
+        if not html:
+            self.logger.error(f"No HTML content for keyword '{keyword}'")
+            return NaverSerpResult(keyword=keyword)
+
+        sections = self.detect_sections(html)
+        above_fold = self.analyze_section_priority(sections)
+        brand_present, brand_name = self.check_brand_zone(html)
+
+        # Build section order
+        section_order = [s.section_type for s in sorted(sections, key=lambda x: x.position)]
+
+        # Count ads
+        ad_sections = [s for s in sections if s.section_type == "ad"]
+        ad_count = sum(s.item_count for s in ad_sections) if ad_sections else 0
+
+        # Check special sections
+        has_view = any(s.section_type == "view_tab" for s in sections)
+        has_place = any(s.section_type == "place" for s in sections)
+        dominant = self._find_dominant_section(sections)
+
+        result = NaverSerpResult(
+            keyword=keyword,
+            sections=sections,
+            section_order=section_order,
+            brand_zone_present=brand_present,
+            brand_zone_brand=brand_name,
+            total_sections=len(sections),
+            above_fold_sections=above_fold,
+            ad_count=ad_count,
+            dominant_section=dominant,
+            has_view_tab=has_view,
+            has_place_section=has_place,
+        )
+        return result
+
+
+# ---------------------------------------------------------------------------
+# Output Helpers
+# ---------------------------------------------------------------------------
+
+
+def result_to_dict(result: NaverSerpResult) -> dict[str, Any]:
+    """Convert NaverSerpResult to a JSON-serializable dictionary."""
+    d = asdict(result)
+    # Remove raw HTML snippets from JSON output to keep it clean
+    for section in d.get("sections", []):
+        section.pop("raw_html_snippet", None)
+    return d
+
+
+def print_rich_report(result: NaverSerpResult) -> None:
+    """Print a human-readable report using rich."""
+    console.rule(f"[bold blue]Naver SERP Analysis: {result.keyword}")
+    console.print(f"[dim]Timestamp: {result.timestamp}[/dim]")
+    console.print()
+
+    # Summary
+    summary_table = Table(title="Summary", show_lines=True)
+    summary_table.add_column("Metric", style="cyan")
+    summary_table.add_column("Value", style="green")
+    summary_table.add_row("Total Sections", str(result.total_sections))
+    summary_table.add_row("Ad Count", str(result.ad_count))
+    summary_table.add_row("Brand Zone", "Yes" if result.brand_zone_present else "No")
+    if result.brand_zone_brand:
+        summary_table.add_row("Brand Name", result.brand_zone_brand)
+    summary_table.add_row("VIEW Tab", "Yes" if result.has_view_tab else "No")
+    summary_table.add_row("Place Section", "Yes" if result.has_place_section else "No")
+    summary_table.add_row("Dominant Section", result.dominant_section or "N/A")
+    console.print(summary_table)
+    console.print()
+
+    # Section Details
+    if result.sections:
+        section_table = Table(title="Detected Sections", show_lines=True)
+        section_table.add_column("#", style="bold")
+        section_table.add_column("Section", style="cyan")
+        section_table.add_column("Display Name", style="magenta")
+        section_table.add_column("Items", style="green")
+        section_table.add_column("Above Fold", style="yellow")
+        section_table.add_column("More Link", style="dim")
+
+        for s in sorted(result.sections, key=lambda x: x.position):
+            section_table.add_row(
+                str(s.position),
+                s.section_type,
+                s.display_name,
+                str(s.item_count),
+                "Yes" if s.is_above_fold else "No",
+                "Yes" if s.has_more_link else "No",
+            )
+        console.print(section_table)
+        console.print()
+
+    # Above-Fold Sections
+    if result.above_fold_sections:
+        console.print("[bold]Above-Fold Section Order:[/bold]")
+        for i, sec in enumerate(result.above_fold_sections, 1):
+            display = SECTION_DISPLAY_NAMES.get(sec, sec)
+            console.print(f"  {i}. {display} ({sec})")
+        console.print()
+
+    # Section Order
+    if result.section_order:
+        console.print("[bold]Full Section Order:[/bold]")
+        order_str = " -> ".join(
+            SECTION_DISPLAY_NAMES.get(s, s) for s in result.section_order
+        )
+        console.print(f"  {order_str}")
+
+    console.rule()
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="Naver SERP composition analysis",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python naver_serp_analyzer.py --keyword "치과 임플란트" --json
+  python naver_serp_analyzer.py --keywords-file keywords.txt --json
+  python naver_serp_analyzer.py --keyword "치과 임플란트" --output report.json
+        """,
+    )
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument(
+        "--keyword",
+        type=str,
+        help="Single keyword to analyze",
+    )
+    group.add_argument(
+        "--keywords-file",
+        type=str,
+        help="Path to file with one keyword per line",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        dest="json_output",
+        help="Output results as JSON",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        help="Write JSON results to file",
+    )
+    return parser
+
+
+def load_keywords(filepath: str) -> list[str]:
+    """Load keywords from a text file, one per line."""
+    path = Path(filepath)
+    if not path.exists():
+        logger.error(f"Keywords file not found: {filepath}")
+        sys.exit(1)
+    keywords = []
+    with open(path, "r", encoding="utf-8") as fh:
+        for line in fh:
+            kw = line.strip()
+            if kw and not kw.startswith("#"):
+                keywords.append(kw)
+    logger.info(f"Loaded {len(keywords)} keywords from {filepath}")
+    return keywords
+
+
+def main() -> None:
+    parser = build_parser()
+    args = parser.parse_args()
+
+    analyzer = NaverSerpAnalyzer()
+
+    # Collect keywords
+    if args.keyword:
+        keywords = [args.keyword]
+    else:
+        keywords = load_keywords(args.keywords_file)
+
+    if not keywords:
+        logger.error("No keywords to analyze")
+        sys.exit(1)
+
+    results: list[dict[str, Any]] = []
+
+    for kw in keywords:
+        console.print(f"\n[bold]Analyzing Naver SERP:[/bold] {kw}")
+        result = analyzer.analyze(kw)
+
+        if args.json_output or args.output:
+            results.append(result_to_dict(result))
+        else:
+            print_rich_report(result)
+
+    # JSON output
+    if args.json_output:
+        output_data = results[0] if len(results) == 1 else results
+        print(json.dumps(output_data, ensure_ascii=False, indent=2))
+
+    if args.output:
+        output_data = results[0] if len(results) == 1 else results
+        output_path = Path(args.output)
+        with open(output_path, "w", encoding="utf-8") as fh:
+            json.dump(output_data, fh, ensure_ascii=False, indent=2)
+        logger.info(f"Results written to {output_path}")
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/20-seo-serp-analysis/code/scripts/requirements.txt b/custom-skills/20-seo-serp-analysis/code/scripts/requirements.txt
new file mode 100644
index 0000000..11238d6
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/code/scripts/requirements.txt
@@ -0,0 +1,9 @@
+# 20-seo-serp-analysis dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+beautifulsoup4>=4.12.0
+lxml>=5.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/20-seo-serp-analysis/code/scripts/serp_analyzer.py b/custom-skills/20-seo-serp-analysis/code/scripts/serp_analyzer.py
new file mode 100644
index 0000000..f681ed4
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/code/scripts/serp_analyzer.py
@@ -0,0 +1,891 @@
+"""
+SERP Analyzer - Google SERP feature detection and competitor mapping
+====================================================================
+Purpose: Analyze Google SERP features, map competitor positions,
+         classify content types, and score SERP opportunities.
+Python: 3.10+
+
+Usage:
+    python serp_analyzer.py --keyword "치과 임플란트" --country kr --json
+    python serp_analyzer.py --keywords-file keywords.txt --country kr --json
+    python serp_analyzer.py --keyword "dental implant" --output serp_report.json
+"""
+
+import argparse
+import json
+import logging
+import re
+import subprocess
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from pathlib import Path
+from typing import Any
+from urllib.parse import urlparse
+
+from rich.console import Console
+from rich.table import Table
+
+# ---------------------------------------------------------------------------
+# Logging
+# ---------------------------------------------------------------------------
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+logger = logging.getLogger(__name__)
+console = Console()
+
+# ---------------------------------------------------------------------------
+# Data Classes
+# ---------------------------------------------------------------------------
+
+
+@dataclass
+class SerpFeatures:
+    """Tracks presence and count of Google SERP features."""
+
+    featured_snippet: bool = False
+    people_also_ask: bool = False
+    local_pack: bool = False
+    knowledge_panel: bool = False
+    video_carousel: bool = False
+    image_pack: bool = False
+    site_links: bool = False
+    ads_top: int = 0
+    ads_bottom: int = 0
+    shopping: bool = False
+
+    @property
+    def feature_count(self) -> int:
+        """Count of boolean features that are present."""
+        count = 0
+        for f in [
+            self.featured_snippet,
+            self.people_also_ask,
+            self.local_pack,
+            self.knowledge_panel,
+            self.video_carousel,
+            self.image_pack,
+            self.site_links,
+            self.shopping,
+        ]:
+            if f:
+                count += 1
+        return count
+
+    @property
+    def has_ads(self) -> bool:
+        return self.ads_top > 0 or self.ads_bottom > 0
+
+
+@dataclass
+class CompetitorPosition:
+    """A single competitor entry in the SERP."""
+
+    position: int
+    url: str
+    domain: str
+    title: str = ""
+    content_type: str = "unknown"
+    is_featured: bool = False
+    has_sitelinks: bool = False
+    estimated_traffic_share: float = 0.0
+
+
+@dataclass
+class SerpResult:
+    """Complete SERP analysis result for a keyword."""
+
+    keyword: str
+    country: str = "us"
+    search_volume: int = 0
+    keyword_difficulty: float = 0.0
+    cpc: float = 0.0
+    serp_features: SerpFeatures = field(default_factory=SerpFeatures)
+    competitors: list[CompetitorPosition] = field(default_factory=list)
+    opportunity_score: int = 0
+    intent_signals: str = "informational"
+    content_type_distribution: dict[str, int] = field(default_factory=dict)
+    volatility: str = "stable"
+    timestamp: str = ""
+
+    def __post_init__(self):
+        if not self.timestamp:
+            self.timestamp = datetime.now().isoformat()
+
+
+# ---------------------------------------------------------------------------
+# Content Type Classifiers
+# ---------------------------------------------------------------------------
+
+# URL path patterns that hint at content type
+URL_CONTENT_PATTERNS: dict[str, list[str]] = {
+    "blog": [
+        r"/blog/",
+        r"/post/",
+        r"/article/",
+        r"/news/",
+        r"/magazine/",
+        r"/journal/",
+        r"/column/",
+        r"/story/",
+        r"\d{4}/\d{2}/",
+    ],
+    "product": [
+        r"/product/",
+        r"/item/",
+        r"/shop/",
+        r"/store/",
+        r"/buy/",
+        r"/p/",
+        r"/goods/",
+        r"/catalog/",
+    ],
+    "service": [
+        r"/service",
+        r"/solution",
+        r"/treatment",
+        r"/procedure",
+        r"/pricing",
+        r"/consultation",
+    ],
+    "news": [
+        r"/news/",
+        r"/press/",
+        r"/media/",
+        r"/release/",
+        r"news\.",
+        r"press\.",
+    ],
+    "video": [
+        r"youtube\.com/watch",
+        r"youtu\.be/",
+        r"vimeo\.com/",
+        r"/video/",
+        r"/watch/",
+    ],
+    "forum": [
+        r"/forum/",
+        r"/community/",
+        r"/discuss",
+        r"/thread/",
+        r"/question/",
+        r"/answers/",
+    ],
+    "wiki": [
+        r"wikipedia\.org",
+        r"/wiki/",
+        r"namu\.wiki",
+    ],
+}
+
+# Title keywords that hint at content type
+TITLE_CONTENT_PATTERNS: dict[str, list[str]] = {
+    "blog": ["블로그", "후기", "리뷰", "review", "guide", "가이드", "팁", "tips"],
+    "product": ["구매", "가격", "buy", "price", "shop", "할인", "sale", "최저가"],
+    "service": ["상담", "치료", "진료", "병원", "클리닉", "clinic", "treatment"],
+    "news": ["뉴스", "속보", "보도", "news", "기사", "report"],
+    "video": ["영상", "동영상", "video", "youtube"],
+    "comparison": ["비교", "vs", "versus", "compare", "차이", "best"],
+}
+
+# CTR distribution by position (approximate click-through rates)
+CTR_BY_POSITION: dict[int, float] = {
+    1: 0.316,
+    2: 0.158,
+    3: 0.110,
+    4: 0.080,
+    5: 0.062,
+    6: 0.049,
+    7: 0.040,
+    8: 0.034,
+    9: 0.029,
+    10: 0.025,
+}
+
+
+# ---------------------------------------------------------------------------
+# SERP Analyzer
+# ---------------------------------------------------------------------------
+
+
+class SerpAnalyzer:
+    """Analyzes Google SERP features, competitor positions, and opportunities."""
+
+    def __init__(self):
+        self.logger = logging.getLogger(self.__class__.__name__)
+
+    # ----- Data Fetching -----
+
+    def get_serp_data(self, keyword: str, country: str = "us") -> dict[str, Any]:
+        """
+        Fetch SERP data via Ahrefs serp-overview MCP tool.
+
+        Uses subprocess to invoke the Ahrefs MCP tool. Falls back to a
+        structured placeholder when the MCP tool is unavailable (e.g., in
+        standalone / CI environments).
+        """
+        self.logger.info(f"Fetching SERP data for '{keyword}' (country={country})")
+
+        try:
+            # Attempt MCP tool call via subprocess
+            cmd = [
+                "claude",
+                "mcp",
+                "call",
+                "ahrefs",
+                "serp-overview",
+                json.dumps({"keyword": keyword, "country": country}),
+            ]
+            result = subprocess.run(
+                cmd,
+                capture_output=True,
+                text=True,
+                timeout=60,
+            )
+            if result.returncode == 0 and result.stdout.strip():
+                data = json.loads(result.stdout)
+                self.logger.info("Successfully fetched SERP data via MCP")
+                return data
+        except (subprocess.TimeoutExpired, FileNotFoundError, json.JSONDecodeError) as exc:
+            self.logger.warning(f"MCP call unavailable ({exc}), using keyword metrics fallback")
+
+        # Fallback: try Ahrefs keywords-explorer-overview
+        try:
+            cmd_kw = [
+                "claude",
+                "mcp",
+                "call",
+                "ahrefs",
+                "keywords-explorer-overview",
+                json.dumps({"keyword": keyword, "country": country}),
+            ]
+            result_kw = subprocess.run(
+                cmd_kw,
+                capture_output=True,
+                text=True,
+                timeout=60,
+            )
+            if result_kw.returncode == 0 and result_kw.stdout.strip():
+                data = json.loads(result_kw.stdout)
+                self.logger.info("Fetched keyword overview via MCP")
+                return data
+        except (subprocess.TimeoutExpired, FileNotFoundError, json.JSONDecodeError) as exc:
+            self.logger.warning(f"Keywords-explorer MCP also unavailable ({exc})")
+
+        # Return empty structure when no MCP tools available
+        self.logger.warning(
+            "No MCP data source available. Run inside Claude Desktop "
+            "or provide data via --input flag."
+        )
+        return {
+            "keyword": keyword,
+            "country": country,
+            "serp": [],
+            "serp_features": {},
+            "metrics": {},
+        }
+
+    # ----- Feature Detection -----
+
+    def detect_features(self, serp_data: dict[str, Any]) -> SerpFeatures:
+        """
+        Identify SERP features from Ahrefs response data.
+
+        Handles both the structured 'serp_features' dict returned by
+        keywords-explorer-overview and the raw SERP items list from
+        serp-overview.
+        """
+        features = SerpFeatures()
+
+        # -- Method 1: structured serp_features from Ahrefs --
+        sf = serp_data.get("serp_features", {})
+        if isinstance(sf, dict):
+            features.featured_snippet = sf.get("featured_snippet", False)
+            features.people_also_ask = sf.get("people_also_ask", False)
+            features.local_pack = sf.get("local_pack", False)
+            features.knowledge_panel = sf.get("knowledge_panel", False) or sf.get(
+                "knowledge_graph", False
+            )
+            features.video_carousel = sf.get("video", False) or sf.get(
+                "video_carousel", False
+            )
+            features.image_pack = sf.get("image_pack", False) or sf.get(
+                "images", False
+            )
+            features.site_links = sf.get("sitelinks", False) or sf.get(
+                "site_links", False
+            )
+            features.shopping = sf.get("shopping_results", False) or sf.get(
+                "shopping", False
+            )
+            features.ads_top = int(sf.get("ads_top", 0) or 0)
+            features.ads_bottom = int(sf.get("ads_bottom", 0) or 0)
+
+        # -- Method 2: infer from raw SERP items list --
+        serp_items = serp_data.get("serp", [])
+        if isinstance(serp_items, list):
+            for item in serp_items:
+                item_type = str(item.get("type", "")).lower()
+                if "featured_snippet" in item_type or item.get("is_featured"):
+                    features.featured_snippet = True
+                if "people_also_ask" in item_type or "paa" in item_type:
+                    features.people_also_ask = True
+                if "local" in item_type or "map" in item_type:
+                    features.local_pack = True
+                if "knowledge" in item_type:
+                    features.knowledge_panel = True
+                if "video" in item_type:
+                    features.video_carousel = True
+                if "image" in item_type:
+                    features.image_pack = True
+                if item.get("sitelinks"):
+                    features.site_links = True
+                if "shopping" in item_type:
+                    features.shopping = True
+                if "ad" in item_type:
+                    pos = item.get("position", 0)
+                    if pos <= 4:
+                        features.ads_top += 1
+                    else:
+                        features.ads_bottom += 1
+
+        return features
+
+    # ----- Competitor Mapping -----
+
+    def map_competitors(self, serp_data: dict[str, Any]) -> list[CompetitorPosition]:
+        """Extract competitor positions and domains from SERP data."""
+        competitors: list[CompetitorPosition] = []
+        serp_items = serp_data.get("serp", [])
+
+        if not isinstance(serp_items, list):
+            return competitors
+
+        for item in serp_items:
+            url = item.get("url", "")
+            if not url:
+                continue
+
+            # Skip ads for organic mapping
+            item_type = str(item.get("type", "")).lower()
+            if "ad" in item_type:
+                continue
+
+            parsed = urlparse(url)
+            domain = parsed.netloc.replace("www.", "")
+            position = int(item.get("position", len(competitors) + 1))
+            title = item.get("title", "")
+
+            content_type = self.classify_content_type(item)
+            traffic_share = CTR_BY_POSITION.get(position, 0.01)
+
+            comp = CompetitorPosition(
+                position=position,
+                url=url,
+                domain=domain,
+                title=title,
+                content_type=content_type,
+                is_featured=bool(item.get("is_featured")),
+                has_sitelinks=bool(item.get("sitelinks")),
+                estimated_traffic_share=round(traffic_share, 4),
+            )
+            competitors.append(comp)
+
+        # Sort by position
+        competitors.sort(key=lambda c: c.position)
+        return competitors
+
+    # ----- Content Type Classification -----
+
+    def classify_content_type(self, result: dict[str, Any]) -> str:
+        """
+        Classify a SERP result as blog/product/service/news/video/forum/wiki
+        based on URL patterns and title keywords.
+        """
+        url = result.get("url", "").lower()
+        title = result.get("title", "").lower()
+
+        scores: dict[str, int] = {}
+
+        # Score from URL patterns
+        for ctype, patterns in URL_CONTENT_PATTERNS.items():
+            for pattern in patterns:
+                if re.search(pattern, url):
+                    scores[ctype] = scores.get(ctype, 0) + 2
+                    break
+
+        # Score from title patterns
+        for ctype, keywords in TITLE_CONTENT_PATTERNS.items():
+            for kw in keywords:
+                if kw.lower() in title:
+                    scores[ctype] = scores.get(ctype, 0) + 1
+
+        if not scores:
+            # Heuristic: if domain is a known authority site
+            parsed = urlparse(url)
+            domain = parsed.netloc.lower()
+            if any(d in domain for d in ["wikipedia", "namu.wiki", "나무위키"]):
+                return "wiki"
+            if any(d in domain for d in ["youtube", "vimeo"]):
+                return "video"
+            if any(d in domain for d in ["naver.com", "tistory.com", "brunch.co.kr"]):
+                return "blog"
+            return "service_page"
+
+        # Return highest scoring type
+        return max(scores, key=scores.get)  # type: ignore[arg-type]
+
+    # ----- Opportunity Scoring -----
+
+    def calculate_opportunity_score(
+        self,
+        features: SerpFeatures,
+        positions: list[CompetitorPosition],
+    ) -> int:
+        """
+        Score SERP opportunity from 0-100.
+
+        Higher scores indicate better opportunity to rank or gain features.
+
+        Factors (additive):
+        - Featured snippet available but could be captured     +15
+        - PAA present (related question opportunity)           +10
+        - No knowledge panel (less SERP real-estate taken)     +10
+        - Low ad count (more organic visibility)               +10
+        - Few sitelinks in top results                         +5
+        - Content diversity (various domains in top 10)        +10
+        - No video carousel (opportunity to add video)         +5
+        - Top results are blogs (easier to outrank)            +10
+        - Image pack absent (image SEO opportunity)            +5
+        - Shopping absent for commercial keywords              +5
+        - Top positions lacking schema/rich results            +5
+
+        Penalty factors (subtractive):
+        - Knowledge panel dominates                            -15
+        - Heavy ad presence (4+ top ads)                       -10
+        - Single domain dominates top 5                        -10
+        """
+        score = 50  # Base score
+
+        # -- Positive signals --
+        if features.featured_snippet:
+            score += 15
+        if features.people_also_ask:
+            score += 10
+        if not features.knowledge_panel:
+            score += 10
+        if features.ads_top <= 1:
+            score += 10
+        elif features.ads_top <= 2:
+            score += 5
+        if not features.video_carousel:
+            score += 5
+        if not features.image_pack:
+            score += 5
+        if not features.shopping:
+            score += 5
+
+        # Domain diversity in top 10
+        if positions:
+            top10_domains = {p.domain for p in positions[:10]}
+            if len(top10_domains) >= 8:
+                score += 10
+            elif len(top10_domains) >= 5:
+                score += 5
+
+            # Blog-heavy top results (easier to compete)
+            blog_count = sum(
+                1 for p in positions[:5] if p.content_type == "blog"
+            )
+            if blog_count >= 3:
+                score += 10
+            elif blog_count >= 2:
+                score += 5
+
+            # Sitelinks reduce available space
+            sitelink_count = sum(1 for p in positions[:5] if p.has_sitelinks)
+            if sitelink_count <= 1:
+                score += 5
+
+            # Single domain dominance penalty
+            domain_counts: dict[str, int] = {}
+            for p in positions[:5]:
+                domain_counts[p.domain] = domain_counts.get(p.domain, 0) + 1
+            if any(c >= 3 for c in domain_counts.values()):
+                score -= 10
+
+        # -- Negative signals --
+        if features.knowledge_panel:
+            score -= 15
+        if features.ads_top >= 4:
+            score -= 10
+        elif features.ads_top >= 3:
+            score -= 5
+
+        # Clamp to 0-100
+        return max(0, min(100, score))
+
+    # ----- Intent Validation -----
+
+    def validate_intent(
+        self,
+        features: SerpFeatures,
+        positions: list[CompetitorPosition],
+    ) -> str:
+        """
+        Infer search intent from SERP composition.
+
+        Returns one of: informational, navigational, commercial, transactional, local
+        """
+        signals: dict[str, int] = {
+            "informational": 0,
+            "navigational": 0,
+            "commercial": 0,
+            "transactional": 0,
+            "local": 0,
+        }
+
+        # Feature-based signals
+        if features.featured_snippet:
+            signals["informational"] += 3
+        if features.people_also_ask:
+            signals["informational"] += 2
+        if features.knowledge_panel:
+            signals["informational"] += 2
+            signals["navigational"] += 2
+        if features.local_pack:
+            signals["local"] += 5
+        if features.shopping:
+            signals["transactional"] += 4
+        if features.has_ads:
+            signals["commercial"] += 2
+            signals["transactional"] += 1
+        if features.ads_top >= 3:
+            signals["transactional"] += 2
+        if features.image_pack:
+            signals["informational"] += 1
+        if features.video_carousel:
+            signals["informational"] += 1
+
+        # Content type signals from top results
+        for pos in positions[:10]:
+            ct = pos.content_type
+            if ct == "blog":
+                signals["informational"] += 1
+            elif ct == "product":
+                signals["transactional"] += 2
+            elif ct == "service":
+                signals["commercial"] += 1
+            elif ct == "news":
+                signals["informational"] += 1
+            elif ct == "video":
+                signals["informational"] += 1
+            elif ct == "wiki":
+                signals["informational"] += 2
+            elif ct == "forum":
+                signals["informational"] += 1
+            elif ct == "comparison":
+                signals["commercial"] += 2
+
+        # Navigational: single domain dominates top 3
+        if positions:
+            top3_domains = [p.domain for p in positions[:3]]
+            if len(set(top3_domains)) == 1:
+                signals["navigational"] += 5
+
+        # Return highest signal
+        return max(signals, key=signals.get)  # type: ignore[arg-type]
+
+    # ----- Content Type Distribution -----
+
+    def _content_type_distribution(
+        self, positions: list[CompetitorPosition]
+    ) -> dict[str, int]:
+        """Count content types across top organic results."""
+        dist: dict[str, int] = {}
+        for p in positions[:10]:
+            dist[p.content_type] = dist.get(p.content_type, 0) + 1
+        return dict(sorted(dist.items(), key=lambda x: x[1], reverse=True))
+
+    # ----- Volatility Assessment -----
+
+    def _assess_volatility(self, serp_data: dict[str, Any]) -> str:
+        """
+        Assess SERP volatility based on available signals.
+
+        Returns: stable, moderate, volatile
+        """
+        # Check if Ahrefs provides a volatility/movement score
+        metrics = serp_data.get("metrics", {})
+        if isinstance(metrics, dict):
+            volatility_score = metrics.get("serp_volatility", None)
+            if volatility_score is not None:
+                if volatility_score < 3:
+                    return "stable"
+                elif volatility_score < 7:
+                    return "moderate"
+                else:
+                    return "volatile"
+
+        # Heuristic: if many results have recent dates, SERP is more volatile
+        serp_items = serp_data.get("serp", [])
+        if isinstance(serp_items, list) and serp_items:
+            recent_count = 0
+            for item in serp_items[:10]:
+                last_seen = item.get("last_seen", "")
+                if last_seen:
+                    try:
+                        dt = datetime.fromisoformat(last_seen.replace("Z", "+00:00"))
+                        if (datetime.now(dt.tzinfo) - dt).days < 30:
+                            recent_count += 1
+                    except (ValueError, TypeError):
+                        pass
+            if recent_count >= 5:
+                return "volatile"
+            elif recent_count >= 3:
+                return "moderate"
+
+        return "stable"
+
+    # ----- Main Analysis Orchestrator -----
+
+    def analyze(self, keyword: str, country: str = "us") -> SerpResult:
+        """
+        Orchestrate full SERP analysis for a single keyword.
+
+        Steps:
+        1. Fetch SERP data from Ahrefs MCP
+        2. Detect SERP features
+        3. Map competitor positions
+        4. Classify content types
+        5. Calculate opportunity score
+        6. Validate search intent
+        7. Assess volatility
+        """
+        serp_data = self.get_serp_data(keyword, country)
+
+        features = self.detect_features(serp_data)
+        positions = self.map_competitors(serp_data)
+        opportunity = self.calculate_opportunity_score(features, positions)
+        intent = self.validate_intent(features, positions)
+        content_dist = self._content_type_distribution(positions)
+        volatility = self._assess_volatility(serp_data)
+
+        # Extract keyword metrics if available
+        metrics = serp_data.get("metrics", {})
+        search_volume = int(metrics.get("search_volume", 0) or 0)
+        keyword_difficulty = float(metrics.get("keyword_difficulty", 0) or 0)
+        cpc = float(metrics.get("cpc", 0) or 0)
+
+        result = SerpResult(
+            keyword=keyword,
+            country=country,
+            search_volume=search_volume,
+            keyword_difficulty=keyword_difficulty,
+            cpc=cpc,
+            serp_features=features,
+            competitors=positions,
+            opportunity_score=opportunity,
+            intent_signals=intent,
+            content_type_distribution=content_dist,
+            volatility=volatility,
+        )
+        return result
+
+
+# ---------------------------------------------------------------------------
+# Output Helpers
+# ---------------------------------------------------------------------------
+
+
+def result_to_dict(result: SerpResult) -> dict[str, Any]:
+    """Convert SerpResult to a JSON-serializable dictionary."""
+    d = asdict(result)
+    return d
+
+
+def print_rich_report(result: SerpResult) -> None:
+    """Print a human-readable report using rich."""
+    console.rule(f"[bold blue]SERP Analysis: {result.keyword}")
+    console.print(f"[dim]Country: {result.country} | Timestamp: {result.timestamp}[/dim]")
+    console.print()
+
+    # Metrics
+    if result.search_volume or result.keyword_difficulty:
+        metrics_table = Table(title="Keyword Metrics", show_lines=True)
+        metrics_table.add_column("Metric", style="cyan")
+        metrics_table.add_column("Value", style="green")
+        metrics_table.add_row("Search Volume", f"{result.search_volume:,}")
+        metrics_table.add_row("Keyword Difficulty", f"{result.keyword_difficulty:.1f}")
+        metrics_table.add_row("CPC", f"${result.cpc:.2f}")
+        console.print(metrics_table)
+        console.print()
+
+    # SERP Features
+    feat = result.serp_features
+    feat_table = Table(title="SERP Features", show_lines=True)
+    feat_table.add_column("Feature", style="cyan")
+    feat_table.add_column("Present", style="green")
+    feat_table.add_row("Featured Snippet", _bool_icon(feat.featured_snippet))
+    feat_table.add_row("People Also Ask", _bool_icon(feat.people_also_ask))
+    feat_table.add_row("Local Pack", _bool_icon(feat.local_pack))
+    feat_table.add_row("Knowledge Panel", _bool_icon(feat.knowledge_panel))
+    feat_table.add_row("Video Carousel", _bool_icon(feat.video_carousel))
+    feat_table.add_row("Image Pack", _bool_icon(feat.image_pack))
+    feat_table.add_row("Site Links", _bool_icon(feat.site_links))
+    feat_table.add_row("Shopping", _bool_icon(feat.shopping))
+    feat_table.add_row("Ads (top)", str(feat.ads_top))
+    feat_table.add_row("Ads (bottom)", str(feat.ads_bottom))
+    console.print(feat_table)
+    console.print()
+
+    # Competitors
+    if result.competitors:
+        comp_table = Table(title="Top Competitors", show_lines=True)
+        comp_table.add_column("#", style="bold")
+        comp_table.add_column("Domain", style="cyan")
+        comp_table.add_column("Type", style="magenta")
+        comp_table.add_column("CTR Share", style="green")
+        comp_table.add_column("Featured", style="yellow")
+        for c in result.competitors[:10]:
+            comp_table.add_row(
+                str(c.position),
+                c.domain,
+                c.content_type,
+                f"{c.estimated_traffic_share:.1%}",
+                _bool_icon(c.is_featured),
+            )
+        console.print(comp_table)
+        console.print()
+
+    # Content Distribution
+    if result.content_type_distribution:
+        dist_table = Table(title="Content Type Distribution (Top 10)", show_lines=True)
+        dist_table.add_column("Content Type", style="cyan")
+        dist_table.add_column("Count", style="green")
+        for ct, count in result.content_type_distribution.items():
+            dist_table.add_row(ct, str(count))
+        console.print(dist_table)
+        console.print()
+
+    # Summary
+    opp_color = "green" if result.opportunity_score >= 60 else (
+        "yellow" if result.opportunity_score >= 40 else "red"
+    )
+    console.print(f"Opportunity Score: [{opp_color}]{result.opportunity_score}/100[/{opp_color}]")
+    console.print(f"Search Intent: [bold]{result.intent_signals}[/bold]")
+    console.print(f"SERP Volatility: [bold]{result.volatility}[/bold]")
+    console.rule()
+
+
+def _bool_icon(val: bool) -> str:
+    """Return Yes/No string for boolean values."""
+    return "Yes" if val else "No"
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="Google SERP feature detection and competitor mapping",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python serp_analyzer.py --keyword "치과 임플란트" --country kr --json
+  python serp_analyzer.py --keywords-file keywords.txt --country kr --output report.json
+        """,
+    )
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument(
+        "--keyword",
+        type=str,
+        help="Single keyword to analyze",
+    )
+    group.add_argument(
+        "--keywords-file",
+        type=str,
+        help="Path to file with one keyword per line",
+    )
+    parser.add_argument(
+        "--country",
+        type=str,
+        default="us",
+        help="Country code for SERP (default: us)",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        dest="json_output",
+        help="Output results as JSON",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        help="Write JSON results to file",
+    )
+    return parser
+
+
+def load_keywords(filepath: str) -> list[str]:
+    """Load keywords from a text file, one per line."""
+    path = Path(filepath)
+    if not path.exists():
+        logger.error(f"Keywords file not found: {filepath}")
+        sys.exit(1)
+    keywords = []
+    with open(path, "r", encoding="utf-8") as fh:
+        for line in fh:
+            kw = line.strip()
+            if kw and not kw.startswith("#"):
+                keywords.append(kw)
+    logger.info(f"Loaded {len(keywords)} keywords from {filepath}")
+    return keywords
+
+
+def main() -> None:
+    parser = build_parser()
+    args = parser.parse_args()
+
+    analyzer = SerpAnalyzer()
+
+    # Collect keywords
+    if args.keyword:
+        keywords = [args.keyword]
+    else:
+        keywords = load_keywords(args.keywords_file)
+
+    if not keywords:
+        logger.error("No keywords to analyze")
+        sys.exit(1)
+
+    results: list[dict[str, Any]] = []
+
+    for kw in keywords:
+        console.print(f"\n[bold]Analyzing:[/bold] {kw}")
+        result = analyzer.analyze(kw, args.country)
+
+        if args.json_output or args.output:
+            results.append(result_to_dict(result))
+        else:
+            print_rich_report(result)
+
+    # JSON output
+    if args.json_output:
+        output_data = results[0] if len(results) == 1 else results
+        print(json.dumps(output_data, ensure_ascii=False, indent=2))
+
+    if args.output:
+        output_data = results[0] if len(results) == 1 else results
+        output_path = Path(args.output)
+        with open(output_path, "w", encoding="utf-8") as fh:
+            json.dump(output_data, fh, ensure_ascii=False, indent=2)
+        logger.info(f"Results written to {output_path}")
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/20-seo-serp-analysis/desktop/SKILL.md b/custom-skills/20-seo-serp-analysis/desktop/SKILL.md
new file mode 100644
index 0000000..d20a850
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/desktop/SKILL.md
@@ -0,0 +1,121 @@
+---
+name: seo-serp-analysis
+description: |
+  SERP analysis for Google and Naver search results.
+  Triggers: SERP analysis, search results, featured snippet, SERP features, Naver SERP, 검색결과 분석, SERP 분석.
+---
+
+# SEO SERP Analysis
+
+## Purpose
+
+Analyze search engine result page composition for Google and Naver. Detect SERP features (featured snippets, PAA, knowledge panels, local pack, video, ads), map competitor positions, score SERP feature opportunities, and analyze Naver section distribution.
+
+## Core Capabilities
+
+1. **Google SERP Feature Detection** - Identify featured snippets, PAA, knowledge panels, local pack, video carousel, ads, image pack, site links, shopping
+2. **Competitor Position Mapping** - Extract domains, positions, content types for top organic results
+3. **Opportunity Scoring** - Score SERP opportunity (0-100) based on feature landscape and competition
+4. **Search Intent Validation** - Infer intent (informational, navigational, commercial, transactional, local) from SERP composition
+5. **Naver SERP Composition** - Detect sections (blog, cafe, knowledge iN, Smart Store, brand zone, VIEW tab), map section priority, analyze brand zone presence
+
+## MCP Tool Usage
+
+### Ahrefs for SERP Data
+```
+mcp__ahrefs__serp-overview: Get SERP results and features for a keyword
+mcp__ahrefs__keywords-explorer-overview: Get keyword metrics, volume, difficulty, and SERP feature flags
+mcp__ahrefs__site-explorer-organic-keywords: Map competitor keyword positions
+```
+
+### Notion for Report Storage
+```
+mcp__notion__notion-create-pages: Save analysis report to SEO Audit Log database
+mcp__notion__notion-update-page: Update existing report entries
+```
+
+### Web Tools for Naver Analysis
+```
+WebSearch: Discover Naver search trends
+WebFetch: Fetch Naver SERP HTML for section analysis
+```
+
+## Workflow
+
+### 1. Google SERP Analysis
+1. Fetch SERP data via `mcp__ahrefs__serp-overview` for the target keyword and country
+2. Detect SERP features (featured snippet, PAA, local pack, knowledge panel, video, ads, images, shopping)
+3. Map competitor positions from organic results (domain, URL, title, position)
+4. Classify content type for each result (blog, product, service, news, video)
+5. Calculate opportunity score (0-100) based on feature landscape
+6. Validate search intent from SERP composition
+7. Assess SERP volatility
+
+### 2. Naver SERP Analysis
+1. Fetch Naver search page for the target keyword
+2. Detect SERP sections (blog, cafe, knowledge iN, Smart Store, brand zone, VIEW tab, news, encyclopedia)
+3. Map section priority (above-fold order)
+4. Check brand zone presence and extract brand name
+5. Count items per section
+6. Identify dominant content section
+
+### 3. Report Generation
+1. Compile results into structured JSON
+2. Generate Korean-language report
+3. Save to Notion SEO Audit Log database
+
+## Output Format
+
+```json
+{
+  "keyword": "치과 임플란트",
+  "country": "kr",
+  "serp_features": {
+    "featured_snippet": true,
+    "people_also_ask": true,
+    "local_pack": true,
+    "knowledge_panel": false,
+    "video_carousel": false,
+    "ads_top": 3,
+    "ads_bottom": 2
+  },
+  "competitors": [
+    {
+      "position": 1,
+      "url": "https://example.com/page",
+      "domain": "example.com",
+      "title": "...",
+      "content_type": "service_page"
+    }
+  ],
+  "opportunity_score": 72,
+  "intent_signals": "commercial",
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Common SERP Features
+
+| Feature | Impact | Opportunity |
+|---------|--------|-------------|
+| Featured Snippet | High visibility above organic | Optimize content format for snippet capture |
+| People Also Ask | Related question visibility | Create FAQ content targeting PAA |
+| Local Pack | Dominates local intent SERPs | Optimize Google Business Profile |
+| Knowledge Panel | Reduces organic CTR | Focus on brand queries and schema |
+| Video Carousel | Visual SERP real estate | Create video content for keyword |
+| Shopping | Transactional intent signal | Product feed optimization |
+
+## Limitations
+
+- Ahrefs SERP data may have a delay (not real-time)
+- Naver SERP HTML structure changes periodically
+- Brand zone detection depends on HTML class patterns
+- Cannot detect personalized SERP results
+
+## Notion Output (Required)
+
+All audit reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category, Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: SERP-YYYYMMDD-NNN
diff --git a/custom-skills/20-seo-serp-analysis/desktop/skill.yaml b/custom-skills/20-seo-serp-analysis/desktop/skill.yaml
new file mode 100644
index 0000000..db1f6b7
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/desktop/skill.yaml
@@ -0,0 +1,14 @@
+# Skill metadata (extracted from SKILL.md frontmatter)
+
+name: seo-serp-analysis
+description: |
+  SERP analysis for Google and Naver. Triggers: SERP analysis, search results, featured snippet, SERP features, Naver SERP.
+
+# Optional fields
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
+
+# triggers: []  # TODO: Extract from description
diff --git a/custom-skills/20-seo-serp-analysis/desktop/tools/ahrefs.md b/custom-skills/20-seo-serp-analysis/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..5d918e0
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/desktop/tools/ahrefs.md
@@ -0,0 +1,15 @@
+# Ahrefs
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/20-seo-serp-analysis/desktop/tools/notion.md b/custom-skills/20-seo-serp-analysis/desktop/tools/notion.md
new file mode 100644
index 0000000..153e086
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/desktop/tools/notion.md
@@ -0,0 +1,15 @@
+# Notion
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/20-seo-serp-analysis/desktop/tools/websearch.md b/custom-skills/20-seo-serp-analysis/desktop/tools/websearch.md
new file mode 100644
index 0000000..36ef238
--- /dev/null
+++ b/custom-skills/20-seo-serp-analysis/desktop/tools/websearch.md
@@ -0,0 +1,15 @@
+# WebSearch
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/21-seo-position-tracking/code/CLAUDE.md b/custom-skills/21-seo-position-tracking/code/CLAUDE.md
new file mode 100644
index 0000000..2609396
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/code/CLAUDE.md
@@ -0,0 +1,148 @@
+# CLAUDE.md
+
+## Overview
+
+Position tracking tool for monitoring keyword rankings via Ahrefs Rank Tracker. Monitors ranking positions, detects position changes with threshold alerts, calculates visibility scores weighted by search volume, compares against competitors, and segments by brand/non-brand keywords.
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Track positions for a project
+python scripts/position_tracker.py --target https://example.com --json
+
+# Generate ranking report
+python scripts/ranking_reporter.py --target https://example.com --period 30 --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `position_tracker.py` | Monitor keyword ranking positions and detect changes | Position data, change alerts, visibility scores |
+| `ranking_reporter.py` | Generate ranking performance reports with trends | Trend analysis, segment reports, competitor comparison |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## Position Tracker
+
+```bash
+# Get current positions
+python scripts/position_tracker.py --target https://example.com --json
+
+# With change threshold alerts (flag positions that moved +-5 or more)
+python scripts/position_tracker.py --target https://example.com --threshold 5 --json
+
+# Filter by keyword segment
+python scripts/position_tracker.py --target https://example.com --segment brand --json
+
+# Compare with competitors
+python scripts/position_tracker.py --target https://example.com --competitor https://comp1.com --json
+```
+
+**Capabilities**:
+- Current ranking position retrieval via Ahrefs Rank Tracker
+- Position change detection with configurable threshold alerts
+- Visibility score calculation (weighted by search volume)
+- Brand vs non-brand keyword segmentation
+- Competitor rank comparison
+- Keyword segment grouping (by intent, cluster, landing page)
+
+## Ranking Reporter
+
+```bash
+# 30-day ranking report
+python scripts/ranking_reporter.py --target https://example.com --period 30 --json
+
+# Quarterly comparison
+python scripts/ranking_reporter.py --target https://example.com --period 90 --json
+
+# Export with competitor comparison
+python scripts/ranking_reporter.py --target https://example.com --competitor https://comp1.com --period 30 --json
+```
+
+**Capabilities**:
+- Period-over-period ranking trends (improved/declined/stable)
+- Top movers (biggest position gains/losses)
+- Visibility score trend over time
+- Segment-level performance breakdown
+- Competitor overlap and position comparison
+- Average position by keyword group
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `rank-tracker-overview` | Get rank tracking overview for project |
+| `rank-tracker-competitors-overview` | Compare against competitors |
+| `rank-tracker-competitors-pages` | Competitor page-level rankings |
+| `rank-tracker-competitors-stats` | Competitor ranking statistics |
+| `rank-tracker-serp-overview` | SERP details for tracked keywords |
+| `management-projects` | List Ahrefs projects |
+| `management-project-keywords` | Get tracked keywords for project |
+
+## Output Format
+
+```json
+{
+  "target": "https://example.com",
+  "total_keywords": 250,
+  "visibility_score": 68.5,
+  "positions": {
+    "top3": 15,
+    "top10": 48,
+    "top20": 92,
+    "top50": 180,
+    "top100": 230
+  },
+  "changes": {
+    "improved": 45,
+    "declined": 30,
+    "stable": 155,
+    "new": 12,
+    "lost": 8
+  },
+  "alerts": [
+    {
+      "keyword": "치과 임플란트 가격",
+      "old_position": 5,
+      "new_position": 15,
+      "change": -10,
+      "volume": 5400
+    }
+  ],
+  "segments": {
+    "brand": {"keywords": 30, "avg_position": 2.1},
+    "non_brand": {"keywords": 220, "avg_position": 24.5}
+  },
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Tracked website URL |
+| Category | Select | Position Tracking |
+| Priority | Select | Based on visibility trend |
+| Found Date | Date | Tracking date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: RANK-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., Visibility Score, SERP, Rank Tracker)
+- URLs and code remain unchanged
diff --git a/custom-skills/21-seo-position-tracking/code/scripts/base_client.py b/custom-skills/21-seo-position-tracking/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/21-seo-position-tracking/code/scripts/position_tracker.py b/custom-skills/21-seo-position-tracking/code/scripts/position_tracker.py
new file mode 100644
index 0000000..d5154d6
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/code/scripts/position_tracker.py
@@ -0,0 +1,786 @@
+"""
+Position Tracker - Keyword Ranking Monitor via Ahrefs Rank Tracker
+==================================================================
+Purpose: Monitor keyword positions, detect changes, calculate visibility scores
+Python: 3.10+
+
+Usage:
+    python position_tracker.py --target https://example.com --json
+    python position_tracker.py --target https://example.com --threshold 5 --json
+    python position_tracker.py --target https://example.com --segment brand --json
+    python position_tracker.py --target https://example.com --competitor https://comp1.com --json
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import math
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from typing import Optional
+from urllib.parse import urlparse
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+# ---------------------------------------------------------------------------
+# CTR curve weights for visibility score (position 1-100)
+# Based on industry-standard organic CTR curves
+# ---------------------------------------------------------------------------
+CTR_WEIGHTS: dict[int, float] = {
+    1: 0.300,
+    2: 0.150,
+    3: 0.100,
+    4: 0.070,
+    5: 0.050,
+    6: 0.038,
+    7: 0.030,
+    8: 0.025,
+    9: 0.020,
+    10: 0.018,
+}
+
+# Positions 11-20 get diminishing CTR
+for _p in range(11, 21):
+    CTR_WEIGHTS[_p] = round(0.015 - (_p - 11) * 0.001, 4)
+
+# Positions 21-50 get minimal CTR
+for _p in range(21, 51):
+    CTR_WEIGHTS[_p] = round(max(0.005 - (_p - 21) * 0.0001, 0.001), 4)
+
+# Positions 51-100 get near-zero CTR
+for _p in range(51, 101):
+    CTR_WEIGHTS[_p] = 0.0005
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+@dataclass
+class KeywordPosition:
+    """Single keyword ranking position."""
+    keyword: str
+    position: int
+    previous_position: Optional[int] = None
+    change: int = 0
+    volume: int = 0
+    url: str = ""
+    intent: str = "informational"
+    is_brand: bool = False
+
+    def __post_init__(self):
+        if self.previous_position is not None:
+            self.change = self.previous_position - self.position
+
+
+@dataclass
+class VisibilityScore:
+    """Weighted visibility score based on CTR curve."""
+    score: float = 0.0
+    top3: int = 0
+    top10: int = 0
+    top20: int = 0
+    top50: int = 0
+    top100: int = 0
+    total_keywords: int = 0
+
+    @property
+    def distribution(self) -> dict:
+        return {
+            "top3": self.top3,
+            "top10": self.top10,
+            "top20": self.top20,
+            "top50": self.top50,
+            "top100": self.top100,
+        }
+
+
+@dataclass
+class PositionAlert:
+    """Alert for significant position change."""
+    keyword: str
+    old_position: int
+    new_position: int
+    change: int
+    volume: int = 0
+    severity: str = "medium"
+
+    def __post_init__(self):
+        abs_change = abs(self.change)
+        if abs_change >= 20:
+            self.severity = "critical"
+        elif abs_change >= 10:
+            self.severity = "high"
+        elif abs_change >= 5:
+            self.severity = "medium"
+        else:
+            self.severity = "low"
+
+
+@dataclass
+class CompetitorComparison:
+    """Competitor ranking comparison result."""
+    competitor: str
+    overlap_keywords: int = 0
+    competitor_better: int = 0
+    target_better: int = 0
+    avg_position_gap: float = 0.0
+    top_gaps: list = field(default_factory=list)
+
+
+@dataclass
+class SegmentData:
+    """Keyword segment aggregation."""
+    name: str
+    keywords: int = 0
+    avg_position: float = 0.0
+    visibility: float = 0.0
+    improved: int = 0
+    declined: int = 0
+    stable: int = 0
+
+
+@dataclass
+class TrackingResult:
+    """Complete position tracking result."""
+    target: str
+    total_keywords: int = 0
+    visibility_score: float = 0.0
+    visibility: Optional[VisibilityScore] = None
+    positions: list[KeywordPosition] = field(default_factory=list)
+    changes: dict = field(default_factory=lambda: {
+        "improved": 0, "declined": 0, "stable": 0, "new": 0, "lost": 0,
+    })
+    alerts: list[PositionAlert] = field(default_factory=list)
+    segments: dict[str, SegmentData] = field(default_factory=dict)
+    competitors: list[CompetitorComparison] = field(default_factory=list)
+    timestamp: str = ""
+
+    def __post_init__(self):
+        if not self.timestamp:
+            self.timestamp = datetime.now().isoformat()
+
+    def to_dict(self) -> dict:
+        """Convert to JSON-serializable dictionary."""
+        result = {
+            "target": self.target,
+            "total_keywords": self.total_keywords,
+            "visibility_score": round(self.visibility_score, 2),
+            "positions": self.visibility.distribution if self.visibility else {},
+            "changes": self.changes,
+            "alerts": [asdict(a) for a in self.alerts],
+            "segments": {
+                k: asdict(v) for k, v in self.segments.items()
+            },
+            "competitors": [asdict(c) for c in self.competitors],
+            "keyword_details": [asdict(p) for p in self.positions],
+            "timestamp": self.timestamp,
+        }
+        return result
+
+
+# ---------------------------------------------------------------------------
+# Position Tracker
+# ---------------------------------------------------------------------------
+class PositionTracker(BaseAsyncClient):
+    """Track keyword ranking positions via Ahrefs Rank Tracker."""
+
+    def __init__(self):
+        super().__init__(
+            max_concurrent=5,
+            requests_per_second=2.0,
+            logger=logger,
+        )
+        self.brand_terms: list[str] = []
+
+    def _extract_domain_brand(self, target: str) -> list[str]:
+        """Extract brand terms from the target domain name."""
+        parsed = urlparse(target)
+        hostname = parsed.hostname or target
+        # Remove TLD and www prefix
+        parts = hostname.replace("www.", "").split(".")
+        brand_parts = []
+        for part in parts:
+            if part not in ("com", "co", "kr", "net", "org", "io", "ai", "www"):
+                brand_parts.append(part.lower())
+                # Also split camelCase and hyphenated forms
+                if "-" in part:
+                    brand_parts.extend(part.lower().split("-"))
+        return list(set(brand_parts))
+
+    async def get_project_keywords(self, target: str) -> list[dict]:
+        """
+        Fetch tracked keywords from Ahrefs management-project-keywords.
+
+        Uses Ahrefs MCP tool: management-project-keywords
+        Returns list of keyword dicts with keyword, volume, intent info.
+        """
+        logger.info(f"Fetching project keywords for: {target}")
+
+        # Step 1: Get project list to find matching project
+        projects = await self._call_ahrefs_projects(target)
+        if not projects:
+            logger.warning(f"No Ahrefs project found for {target}. Using rank-tracker-overview directly.")
+            return []
+
+        project_id = projects[0].get("id", "")
+
+        # Step 2: Fetch keywords for the project
+        keywords_data = await self._call_ahrefs_project_keywords(project_id)
+        return keywords_data
+
+    async def _call_ahrefs_projects(self, target: str) -> list[dict]:
+        """
+        Call Ahrefs management-projects MCP tool.
+        In production, this calls the MCP tool. For standalone, reads from config/cache.
+        """
+        # Simulated MCP call structure - in production this calls:
+        # mcp__ahrefs__management-projects
+        logger.info("Calling Ahrefs management-projects...")
+        try:
+            import subprocess
+            result = subprocess.run(
+                ["mcp-cli", "call", "ahrefs/management-projects", json.dumps({})],
+                capture_output=True, text=True, timeout=30,
+            )
+            if result.returncode == 0:
+                return json.loads(result.stdout).get("projects", [])
+        except (FileNotFoundError, subprocess.TimeoutExpired, json.JSONDecodeError):
+            pass
+        # Return empty if MCP not available - caller handles gracefully
+        return []
+
+    async def _call_ahrefs_project_keywords(self, project_id: str) -> list[dict]:
+        """
+        Call Ahrefs management-project-keywords MCP tool.
+        """
+        logger.info(f"Calling Ahrefs management-project-keywords for project: {project_id}")
+        try:
+            import subprocess
+            result = subprocess.run(
+                ["mcp-cli", "call", "ahrefs/management-project-keywords",
+                 json.dumps({"project_id": project_id})],
+                capture_output=True, text=True, timeout=30,
+            )
+            if result.returncode == 0:
+                return json.loads(result.stdout).get("keywords", [])
+        except (FileNotFoundError, subprocess.TimeoutExpired, json.JSONDecodeError):
+            pass
+        return []
+
+    async def get_current_positions(self, target: str) -> list[KeywordPosition]:
+        """
+        Fetch current keyword positions via Ahrefs rank-tracker-overview.
+
+        Returns list of KeywordPosition objects with current and previous positions.
+        """
+        logger.info(f"Fetching current positions for: {target}")
+        self.brand_terms = self._extract_domain_brand(target)
+
+        raw_data = await self._call_rank_tracker_overview(target)
+        positions: list[KeywordPosition] = []
+
+        for item in raw_data:
+            keyword = item.get("keyword", "")
+            current_pos = item.get("position", 0)
+            prev_pos = item.get("previous_position")
+            volume = item.get("volume", 0)
+            url = item.get("url", "")
+            intent = item.get("intent", "informational")
+
+            # Determine if brand keyword
+            is_brand = self._is_brand_keyword(keyword)
+
+            kp = KeywordPosition(
+                keyword=keyword,
+                position=current_pos,
+                previous_position=prev_pos,
+                volume=volume,
+                url=url,
+                intent=intent,
+                is_brand=is_brand,
+            )
+            positions.append(kp)
+
+        logger.info(f"Retrieved {len(positions)} keyword positions")
+        return positions
+
+    async def _call_rank_tracker_overview(self, target: str) -> list[dict]:
+        """
+        Call Ahrefs rank-tracker-overview MCP tool.
+        """
+        logger.info(f"Calling Ahrefs rank-tracker-overview for: {target}")
+        try:
+            import subprocess
+            result = subprocess.run(
+                ["mcp-cli", "call", "ahrefs/rank-tracker-overview",
+                 json.dumps({"target": target})],
+                capture_output=True, text=True, timeout=60,
+            )
+            if result.returncode == 0:
+                data = json.loads(result.stdout)
+                return data.get("keywords", data.get("results", []))
+        except (FileNotFoundError, subprocess.TimeoutExpired, json.JSONDecodeError):
+            pass
+        return []
+
+    def _is_brand_keyword(self, keyword: str) -> bool:
+        """Check if a keyword is brand-related based on domain name."""
+        keyword_lower = keyword.lower()
+        for term in self.brand_terms:
+            if term in keyword_lower:
+                return True
+        return False
+
+    def detect_changes(
+        self,
+        positions: list[KeywordPosition],
+        threshold: int = 3,
+    ) -> tuple[dict, list[PositionAlert]]:
+        """
+        Detect significant position changes and generate alerts.
+
+        Args:
+            positions: List of current keyword positions with previous data
+            threshold: Minimum position change to trigger an alert
+
+        Returns:
+            Tuple of (change_summary_dict, list_of_alerts)
+        """
+        changes = {
+            "improved": 0,
+            "declined": 0,
+            "stable": 0,
+            "new": 0,
+            "lost": 0,
+        }
+        alerts: list[PositionAlert] = []
+
+        for kp in positions:
+            if kp.previous_position is None:
+                changes["new"] += 1
+                continue
+
+            if kp.position == 0 and kp.previous_position > 0:
+                changes["lost"] += 1
+                alert = PositionAlert(
+                    keyword=kp.keyword,
+                    old_position=kp.previous_position,
+                    new_position=0,
+                    change=-kp.previous_position,
+                    volume=kp.volume,
+                )
+                alerts.append(alert)
+                continue
+
+            change = kp.change  # positive = improved, negative = declined
+            if change > 0:
+                changes["improved"] += 1
+            elif change < 0:
+                changes["declined"] += 1
+            else:
+                changes["stable"] += 1
+
+            # Generate alert if change exceeds threshold
+            if abs(change) >= threshold:
+                alert = PositionAlert(
+                    keyword=kp.keyword,
+                    old_position=kp.previous_position,
+                    new_position=kp.position,
+                    change=change,
+                    volume=kp.volume,
+                )
+                alerts.append(alert)
+
+        # Sort alerts by severity (critical first) then by volume (high first)
+        severity_order = {"critical": 0, "high": 1, "medium": 2, "low": 3}
+        alerts.sort(key=lambda a: (severity_order.get(a.severity, 4), -a.volume))
+
+        logger.info(
+            f"Changes detected - improved: {changes['improved']}, "
+            f"declined: {changes['declined']}, stable: {changes['stable']}, "
+            f"new: {changes['new']}, lost: {changes['lost']}"
+        )
+        logger.info(f"Alerts generated: {len(alerts)} (threshold: {threshold})")
+
+        return changes, alerts
+
+    def calculate_visibility(self, positions: list[KeywordPosition]) -> VisibilityScore:
+        """
+        Calculate weighted visibility score based on CTR curve.
+
+        Visibility = sum(keyword_volume * ctr_weight_for_position) / sum(keyword_volume)
+        Score normalized to 0-100 scale.
+        """
+        vis = VisibilityScore()
+        total_weighted = 0.0
+        total_volume = 0
+
+        for kp in positions:
+            if kp.position <= 0 or kp.position > 100:
+                continue
+
+            vis.total_keywords += 1
+            volume = max(kp.volume, 1)  # Avoid zero volume
+            total_volume += volume
+
+            # Position bucket counting
+            if kp.position <= 3:
+                vis.top3 += 1
+            if kp.position <= 10:
+                vis.top10 += 1
+            if kp.position <= 20:
+                vis.top20 += 1
+            if kp.position <= 50:
+                vis.top50 += 1
+            if kp.position <= 100:
+                vis.top100 += 1
+
+            # Weighted visibility
+            ctr = CTR_WEIGHTS.get(kp.position, 0.0005)
+            total_weighted += volume * ctr
+
+        if total_volume > 0:
+            # Normalize: max possible is if all keywords were position 1
+            max_possible = total_volume * CTR_WEIGHTS[1]
+            vis.score = (total_weighted / max_possible) * 100.0
+        else:
+            vis.score = 0.0
+
+        logger.info(
+            f"Visibility score: {vis.score:.2f} | "
+            f"Top3: {vis.top3}, Top10: {vis.top10}, Top20: {vis.top20}"
+        )
+
+        return vis
+
+    def segment_keywords(
+        self,
+        positions: list[KeywordPosition],
+        filter_segment: Optional[str] = None,
+    ) -> dict[str, SegmentData]:
+        """
+        Segment keywords into brand/non-brand and by intent type.
+
+        Args:
+            positions: List of keyword positions
+            filter_segment: Optional filter - 'brand', 'non_brand', or intent type
+
+        Returns:
+            Dictionary of segment name to SegmentData
+        """
+        segments: dict[str, list[KeywordPosition]] = {
+            "brand": [],
+            "non_brand": [],
+        }
+        intent_segments: dict[str, list[KeywordPosition]] = {}
+
+        for kp in positions:
+            # Brand segmentation
+            if kp.is_brand:
+                segments["brand"].append(kp)
+            else:
+                segments["non_brand"].append(kp)
+
+            # Intent segmentation
+            intent_key = kp.intent.lower() if kp.intent else "informational"
+            if intent_key not in intent_segments:
+                intent_segments[intent_key] = []
+            intent_segments[intent_key].append(kp)
+
+        # Merge intent segments into main segments
+        for intent_key, kps in intent_segments.items():
+            segments[f"intent_{intent_key}"] = kps
+
+        # Calculate segment stats
+        result: dict[str, SegmentData] = {}
+        for seg_name, kps in segments.items():
+            if filter_segment and seg_name != filter_segment:
+                continue
+
+            if not kps:
+                continue
+
+            active_positions = [kp for kp in kps if kp.position > 0]
+            avg_pos = (
+                sum(kp.position for kp in active_positions) / len(active_positions)
+                if active_positions else 0.0
+            )
+
+            vis = self.calculate_visibility(kps)
+
+            improved = sum(1 for kp in kps if kp.change > 0)
+            declined = sum(1 for kp in kps if kp.change < 0)
+            stable = sum(1 for kp in kps if kp.change == 0 and kp.previous_position is not None)
+
+            result[seg_name] = SegmentData(
+                name=seg_name,
+                keywords=len(kps),
+                avg_position=round(avg_pos, 1),
+                visibility=round(vis.score, 2),
+                improved=improved,
+                declined=declined,
+                stable=stable,
+            )
+
+        return result
+
+    async def compare_competitors(
+        self,
+        target: str,
+        competitors: list[str],
+    ) -> list[CompetitorComparison]:
+        """
+        Compare ranking positions against competitors.
+
+        Uses Ahrefs rank-tracker-competitors-overview MCP tool.
+        """
+        comparisons: list[CompetitorComparison] = []
+
+        for competitor in competitors:
+            logger.info(f"Comparing with competitor: {competitor}")
+            comp_data = await self._call_competitors_overview(target, competitor)
+
+            comparison = CompetitorComparison(competitor=competitor)
+
+            if comp_data:
+                comparison.overlap_keywords = comp_data.get("overlap_keywords", 0)
+                comparison.competitor_better = comp_data.get("competitor_better", 0)
+                comparison.target_better = comp_data.get("target_better", 0)
+                comparison.avg_position_gap = comp_data.get("avg_position_gap", 0.0)
+
+                # Extract top gaps (keywords where competitor outranks us most)
+                top_gaps = comp_data.get("top_gaps", [])
+                comparison.top_gaps = top_gaps[:10]
+
+            comparisons.append(comparison)
+
+        return comparisons
+
+    async def _call_competitors_overview(self, target: str, competitor: str) -> dict:
+        """
+        Call Ahrefs rank-tracker-competitors-overview MCP tool.
+        """
+        logger.info(f"Calling Ahrefs rank-tracker-competitors-overview...")
+        try:
+            import subprocess
+            result = subprocess.run(
+                ["mcp-cli", "call", "ahrefs/rank-tracker-competitors-overview",
+                 json.dumps({"target": target, "competitor": competitor})],
+                capture_output=True, text=True, timeout=60,
+            )
+            if result.returncode == 0:
+                return json.loads(result.stdout)
+        except (FileNotFoundError, subprocess.TimeoutExpired, json.JSONDecodeError):
+            pass
+        return {}
+
+    async def analyze(
+        self,
+        target: str,
+        threshold: int = 3,
+        competitors: Optional[list[str]] = None,
+        segment_filter: Optional[str] = None,
+    ) -> TrackingResult:
+        """
+        Orchestrate full position tracking analysis.
+
+        Args:
+            target: Target website URL
+            threshold: Position change threshold for alerts
+            competitors: List of competitor URLs to compare
+            segment_filter: Optional segment filter (brand, non_brand, intent_*)
+
+        Returns:
+            Complete TrackingResult with all analysis data
+        """
+        logger.info(f"Starting position tracking analysis for: {target}")
+        logger.info(f"Threshold: {threshold}, Competitors: {competitors or 'none'}")
+
+        result = TrackingResult(target=target)
+
+        # Step 1: Fetch current positions
+        positions = await self.get_current_positions(target)
+
+        if not positions:
+            logger.warning("No position data retrieved. Check Ahrefs project configuration.")
+            return result
+
+        result.positions = positions
+        result.total_keywords = len(positions)
+
+        # Step 2: Detect changes and generate alerts
+        changes, alerts = self.detect_changes(positions, threshold)
+        result.changes = changes
+        result.alerts = alerts
+
+        # Step 3: Calculate visibility score
+        visibility = self.calculate_visibility(positions)
+        result.visibility = visibility
+        result.visibility_score = visibility.score
+
+        # Step 4: Segment keywords
+        segments = self.segment_keywords(positions, segment_filter)
+        result.segments = segments
+
+        # Step 5: Compare with competitors (if provided)
+        if competitors:
+            comp_results = await self.compare_competitors(target, competitors)
+            result.competitors = comp_results
+
+        logger.info(f"Analysis complete. Total keywords: {result.total_keywords}")
+        logger.info(f"Visibility score: {result.visibility_score:.2f}")
+
+        return result
+
+
+# ---------------------------------------------------------------------------
+# Output formatters
+# ---------------------------------------------------------------------------
+def format_text_report(result: TrackingResult) -> str:
+    """Format tracking result as human-readable text report."""
+    lines = []
+    lines.append("=" * 60)
+    lines.append(f"Position Tracking Report: {result.target}")
+    lines.append(f"Timestamp: {result.timestamp}")
+    lines.append("=" * 60)
+
+    # Visibility overview
+    lines.append(f"\nVisibility Score: {result.visibility_score:.2f}/100")
+    lines.append(f"Total Keywords Tracked: {result.total_keywords}")
+
+    if result.visibility:
+        vis = result.visibility
+        lines.append(f"\nPosition Distribution:")
+        lines.append(f"  Top 3:   {vis.top3}")
+        lines.append(f"  Top 10:  {vis.top10}")
+        lines.append(f"  Top 20:  {vis.top20}")
+        lines.append(f"  Top 50:  {vis.top50}")
+        lines.append(f"  Top 100: {vis.top100}")
+
+    # Changes summary
+    ch = result.changes
+    lines.append(f"\nPosition Changes:")
+    lines.append(f"  Improved: {ch.get('improved', 0)}")
+    lines.append(f"  Declined: {ch.get('declined', 0)}")
+    lines.append(f"  Stable:   {ch.get('stable', 0)}")
+    lines.append(f"  New:      {ch.get('new', 0)}")
+    lines.append(f"  Lost:     {ch.get('lost', 0)}")
+
+    # Alerts
+    if result.alerts:
+        lines.append(f"\nAlerts ({len(result.alerts)}):")
+        lines.append("-" * 60)
+        for alert in result.alerts[:20]:
+            direction = "UP" if alert.change > 0 else "DOWN"
+            lines.append(
+                f"  [{alert.severity.upper()}] {alert.keyword}: "
+                f"{alert.old_position} -> {alert.new_position} "
+                f"({direction} {abs(alert.change)}) | Vol: {alert.volume}"
+            )
+
+    # Segments
+    if result.segments:
+        lines.append(f"\nSegments:")
+        lines.append("-" * 60)
+        for name, seg in result.segments.items():
+            lines.append(
+                f"  {name}: {seg.keywords} keywords, "
+                f"avg pos {seg.avg_position}, "
+                f"vis {seg.visibility}"
+            )
+
+    # Competitors
+    if result.competitors:
+        lines.append(f"\nCompetitor Comparison:")
+        lines.append("-" * 60)
+        for comp in result.competitors:
+            lines.append(f"  vs {comp.competitor}:")
+            lines.append(f"    Overlap: {comp.overlap_keywords} keywords")
+            lines.append(f"    We win: {comp.target_better}")
+            lines.append(f"    They win: {comp.competitor_better}")
+            lines.append(f"    Avg gap: {comp.avg_position_gap:.1f}")
+
+    lines.append("\n" + "=" * 60)
+    return "\n".join(lines)
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Position Tracker - Monitor keyword rankings via Ahrefs Rank Tracker",
+    )
+    parser.add_argument(
+        "--target",
+        required=True,
+        help="Target website URL (e.g., https://example.com)",
+    )
+    parser.add_argument(
+        "--threshold",
+        type=int,
+        default=3,
+        help="Position change threshold for alerts (default: 3)",
+    )
+    parser.add_argument(
+        "--segment",
+        choices=["brand", "non_brand", "intent_informational",
+                 "intent_commercial", "intent_transactional", "intent_navigational"],
+        default=None,
+        help="Filter results by keyword segment",
+    )
+    parser.add_argument(
+        "--competitor",
+        action="append",
+        dest="competitors",
+        default=[],
+        help="Competitor URL to compare (repeatable)",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        dest="json_output",
+        help="Output in JSON format",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        default=None,
+        help="Save output to file path",
+    )
+    return parser.parse_args()
+
+
+async def main():
+    args = parse_args()
+
+    tracker = PositionTracker()
+
+    result = await tracker.analyze(
+        target=args.target,
+        threshold=args.threshold,
+        competitors=args.competitors,
+        segment_filter=args.segment,
+    )
+
+    if args.json_output:
+        output = json.dumps(result.to_dict(), ensure_ascii=False, indent=2)
+    else:
+        output = format_text_report(result)
+
+    if args.output:
+        with open(args.output, "w", encoding="utf-8") as f:
+            f.write(output)
+        logger.info(f"Output saved to: {args.output}")
+    else:
+        print(output)
+
+    tracker.print_stats()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/21-seo-position-tracking/code/scripts/ranking_reporter.py b/custom-skills/21-seo-position-tracking/code/scripts/ranking_reporter.py
new file mode 100644
index 0000000..10f60b8
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/code/scripts/ranking_reporter.py
@@ -0,0 +1,728 @@
+"""
+Ranking Reporter - Ranking Performance Reports with Trends
+==========================================================
+Purpose: Generate ranking reports with trend analysis, top movers, and competitor comparison
+Python: 3.10+
+
+Usage:
+    python ranking_reporter.py --target https://example.com --period 30 --json
+    python ranking_reporter.py --target https://example.com --period 90 --json
+    python ranking_reporter.py --target https://example.com --competitor https://comp1.com --period 30 --json
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime, timedelta
+from typing import Optional
+from urllib.parse import urlparse
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+# CTR weights for impact scoring (same as position_tracker)
+CTR_WEIGHTS: dict[int, float] = {
+    1: 0.300, 2: 0.150, 3: 0.100, 4: 0.070, 5: 0.050,
+    6: 0.038, 7: 0.030, 8: 0.025, 9: 0.020, 10: 0.018,
+}
+for _p in range(11, 21):
+    CTR_WEIGHTS[_p] = round(0.015 - (_p - 11) * 0.001, 4)
+for _p in range(21, 51):
+    CTR_WEIGHTS[_p] = round(max(0.005 - (_p - 21) * 0.0001, 0.001), 4)
+for _p in range(51, 101):
+    CTR_WEIGHTS[_p] = 0.0005
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+@dataclass
+class PositionSnapshot:
+    """A single position measurement at a point in time."""
+    date: str
+    position: int
+    volume: int = 0
+    url: str = ""
+
+
+@dataclass
+class RankingTrend:
+    """Keyword ranking trend over time."""
+    keyword: str
+    positions_over_time: list[PositionSnapshot] = field(default_factory=list)
+    trend_direction: str = "stable"  # improved, declined, stable, new, lost
+    avg_position: float = 0.0
+    current_position: int = 0
+    start_position: int = 0
+    total_change: int = 0
+    volume: int = 0
+    intent: str = "informational"
+    is_brand: bool = False
+
+    def compute_trend(self):
+        """Compute trend direction and average from position history."""
+        if not self.positions_over_time:
+            self.trend_direction = "stable"
+            return
+
+        positions = [s.position for s in self.positions_over_time if s.position > 0]
+        if not positions:
+            self.trend_direction = "lost"
+            return
+
+        self.avg_position = sum(positions) / len(positions)
+        self.current_position = positions[-1]
+        self.start_position = positions[0]
+        self.total_change = self.start_position - self.current_position
+
+        # Determine trend using linear regression direction
+        if len(positions) >= 2:
+            n = len(positions)
+            x_mean = (n - 1) / 2.0
+            y_mean = sum(positions) / n
+            numerator = sum((i - x_mean) * (p - y_mean) for i, p in enumerate(positions))
+            denominator = sum((i - x_mean) ** 2 for i in range(n))
+
+            if denominator > 0:
+                slope = numerator / denominator
+                # Negative slope means position number decreasing = improving
+                if slope < -0.5:
+                    self.trend_direction = "improved"
+                elif slope > 0.5:
+                    self.trend_direction = "declined"
+                else:
+                    self.trend_direction = "stable"
+            else:
+                self.trend_direction = "stable"
+
+        if self.volume == 0 and self.positions_over_time:
+            self.volume = self.positions_over_time[-1].volume
+
+
+@dataclass
+class TopMover:
+    """Keyword with significant position change."""
+    keyword: str
+    position_change: int
+    current_position: int = 0
+    previous_position: int = 0
+    volume: int = 0
+    impact_score: float = 0.0
+    direction: str = "improved"
+
+    def calculate_impact(self):
+        """Calculate impact score: volume * CTR delta."""
+        old_ctr = CTR_WEIGHTS.get(self.previous_position, 0.0005) if self.previous_position > 0 else 0.0
+        new_ctr = CTR_WEIGHTS.get(self.current_position, 0.0005) if self.current_position > 0 else 0.0
+        ctr_delta = abs(new_ctr - old_ctr)
+        self.impact_score = round(self.volume * ctr_delta, 2)
+        self.direction = "improved" if self.position_change > 0 else "declined"
+
+
+@dataclass
+class SegmentReport:
+    """Performance breakdown for a keyword segment."""
+    segment_name: str
+    total_keywords: int = 0
+    avg_position: float = 0.0
+    avg_position_change: float = 0.0
+    visibility_score: float = 0.0
+    improved_count: int = 0
+    declined_count: int = 0
+    stable_count: int = 0
+    top_gainers: list[TopMover] = field(default_factory=list)
+    top_losers: list[TopMover] = field(default_factory=list)
+
+
+@dataclass
+class CompetitorReport:
+    """Competitor comparison for a reporting period."""
+    competitor: str
+    our_visibility: float = 0.0
+    their_visibility: float = 0.0
+    overlap_keywords: int = 0
+    keywords_we_lead: int = 0
+    keywords_they_lead: int = 0
+    notable_gaps: list[dict] = field(default_factory=list)
+
+
+@dataclass
+class RankingReport:
+    """Complete ranking performance report."""
+    target: str
+    period_days: int = 30
+    period_start: str = ""
+    period_end: str = ""
+    total_keywords: int = 0
+    current_visibility: float = 0.0
+    previous_visibility: float = 0.0
+    visibility_change: float = 0.0
+    trend_summary: dict = field(default_factory=lambda: {
+        "improved": 0, "declined": 0, "stable": 0, "new": 0, "lost": 0,
+    })
+    top_gainers: list[TopMover] = field(default_factory=list)
+    top_losers: list[TopMover] = field(default_factory=list)
+    segments: list[SegmentReport] = field(default_factory=list)
+    competitors: list[CompetitorReport] = field(default_factory=list)
+    keyword_trends: list[RankingTrend] = field(default_factory=list)
+    timestamp: str = ""
+
+    def __post_init__(self):
+        if not self.timestamp:
+            self.timestamp = datetime.now().isoformat()
+        if not self.period_end:
+            self.period_end = datetime.now().strftime("%Y-%m-%d")
+        if not self.period_start:
+            start = datetime.now() - timedelta(days=self.period_days)
+            self.period_start = start.strftime("%Y-%m-%d")
+
+    def to_dict(self) -> dict:
+        """Convert to JSON-serializable dictionary."""
+        return {
+            "target": self.target,
+            "period": {
+                "days": self.period_days,
+                "start": self.period_start,
+                "end": self.period_end,
+            },
+            "total_keywords": self.total_keywords,
+            "visibility": {
+                "current": round(self.current_visibility, 2),
+                "previous": round(self.previous_visibility, 2),
+                "change": round(self.visibility_change, 2),
+            },
+            "trend_summary": self.trend_summary,
+            "top_gainers": [asdict(m) for m in self.top_gainers],
+            "top_losers": [asdict(m) for m in self.top_losers],
+            "segments": [asdict(s) for s in self.segments],
+            "competitors": [asdict(c) for c in self.competitors],
+            "keyword_trends": [
+                {
+                    "keyword": t.keyword,
+                    "trend_direction": t.trend_direction,
+                    "avg_position": round(t.avg_position, 1),
+                    "current_position": t.current_position,
+                    "start_position": t.start_position,
+                    "total_change": t.total_change,
+                    "volume": t.volume,
+                }
+                for t in self.keyword_trends
+            ],
+            "timestamp": self.timestamp,
+        }
+
+
+# ---------------------------------------------------------------------------
+# Ranking Reporter
+# ---------------------------------------------------------------------------
+class RankingReporter(BaseAsyncClient):
+    """Generate ranking performance reports with trend analysis."""
+
+    def __init__(self):
+        super().__init__(
+            max_concurrent=5,
+            requests_per_second=2.0,
+            logger=logger,
+        )
+
+    def _extract_domain_brand(self, target: str) -> list[str]:
+        """Extract brand terms from the target domain name."""
+        parsed = urlparse(target)
+        hostname = parsed.hostname or target
+        parts = hostname.replace("www.", "").split(".")
+        brand_parts = []
+        for part in parts:
+            if part not in ("com", "co", "kr", "net", "org", "io", "ai", "www"):
+                brand_parts.append(part.lower())
+                if "-" in part:
+                    brand_parts.extend(part.lower().split("-"))
+        return list(set(brand_parts))
+
+    async def get_historical_positions(
+        self,
+        target: str,
+        period_days: int = 30,
+    ) -> list[RankingTrend]:
+        """
+        Fetch historical position data from Ahrefs rank-tracker-overview
+        with date range parameters.
+
+        Returns list of RankingTrend objects with position snapshots over time.
+        """
+        logger.info(f"Fetching historical positions for {target} ({period_days} days)")
+        brand_terms = self._extract_domain_brand(target)
+
+        end_date = datetime.now().strftime("%Y-%m-%d")
+        start_date = (datetime.now() - timedelta(days=period_days)).strftime("%Y-%m-%d")
+
+        raw_data = await self._call_rank_tracker_historical(target, start_date, end_date)
+
+        trends: dict[str, RankingTrend] = {}
+        for item in raw_data:
+            keyword = item.get("keyword", "")
+            if keyword not in trends:
+                is_brand = any(term in keyword.lower() for term in brand_terms)
+                trends[keyword] = RankingTrend(
+                    keyword=keyword,
+                    volume=item.get("volume", 0),
+                    intent=item.get("intent", "informational"),
+                    is_brand=is_brand,
+                )
+
+            snapshot = PositionSnapshot(
+                date=item.get("date", end_date),
+                position=item.get("position", 0),
+                volume=item.get("volume", 0),
+                url=item.get("url", ""),
+            )
+            trends[keyword].positions_over_time.append(snapshot)
+
+        # Sort snapshots by date and compute trends
+        for trend in trends.values():
+            trend.positions_over_time.sort(key=lambda s: s.date)
+            trend.compute_trend()
+
+        logger.info(f"Retrieved trends for {len(trends)} keywords")
+        return list(trends.values())
+
+    async def _call_rank_tracker_historical(
+        self, target: str, start_date: str, end_date: str,
+    ) -> list[dict]:
+        """Call Ahrefs rank-tracker-overview with date range."""
+        logger.info(f"Calling Ahrefs rank-tracker-overview ({start_date} to {end_date})...")
+        try:
+            import subprocess
+            result = subprocess.run(
+                ["mcp-cli", "call", "ahrefs/rank-tracker-overview",
+                 json.dumps({
+                     "target": target,
+                     "date_from": start_date,
+                     "date_to": end_date,
+                 })],
+                capture_output=True, text=True, timeout=60,
+            )
+            if result.returncode == 0:
+                data = json.loads(result.stdout)
+                return data.get("keywords", data.get("results", []))
+        except (FileNotFoundError, subprocess.TimeoutExpired, json.JSONDecodeError):
+            pass
+        return []
+
+    def calculate_trends(self, trends: list[RankingTrend]) -> dict:
+        """
+        Compute overall trend summary from keyword trends.
+
+        Returns dict with improved/declined/stable/new/lost counts.
+        """
+        summary = {
+            "improved": 0,
+            "declined": 0,
+            "stable": 0,
+            "new": 0,
+            "lost": 0,
+        }
+        for trend in trends:
+            direction = trend.trend_direction
+            if direction in summary:
+                summary[direction] += 1
+            else:
+                summary["stable"] += 1
+
+        logger.info(
+            f"Trend summary: improved={summary['improved']}, "
+            f"declined={summary['declined']}, stable={summary['stable']}"
+        )
+        return summary
+
+    def find_top_movers(
+        self,
+        trends: list[RankingTrend],
+        limit: int = 10,
+    ) -> tuple[list[TopMover], list[TopMover]]:
+        """
+        Find keywords with biggest position gains and losses.
+
+        Returns tuple of (top_gainers, top_losers) sorted by impact score.
+        """
+        gainers: list[TopMover] = []
+        losers: list[TopMover] = []
+
+        for trend in trends:
+            if not trend.positions_over_time or len(trend.positions_over_time) < 2:
+                continue
+
+            first_pos = trend.start_position
+            last_pos = trend.current_position
+
+            if first_pos <= 0 or last_pos <= 0:
+                continue
+
+            change = first_pos - last_pos  # positive = improved
+
+            mover = TopMover(
+                keyword=trend.keyword,
+                position_change=change,
+                current_position=last_pos,
+                previous_position=first_pos,
+                volume=trend.volume,
+            )
+            mover.calculate_impact()
+
+            if change > 0:
+                gainers.append(mover)
+            elif change < 0:
+                losers.append(mover)
+
+        # Sort by impact score descending
+        gainers.sort(key=lambda m: m.impact_score, reverse=True)
+        losers.sort(key=lambda m: m.impact_score, reverse=True)
+
+        logger.info(f"Top movers: {len(gainers)} gainers, {len(losers)} losers")
+        return gainers[:limit], losers[:limit]
+
+    def _calculate_visibility_score(self, trends: list[RankingTrend], use_start: bool = False) -> float:
+        """Calculate visibility score from trends (current or start positions)."""
+        total_weighted = 0.0
+        total_volume = 0
+
+        for trend in trends:
+            pos = trend.start_position if use_start else trend.current_position
+            if pos <= 0 or pos > 100:
+                continue
+            volume = max(trend.volume, 1)
+            total_volume += volume
+            ctr = CTR_WEIGHTS.get(pos, 0.0005)
+            total_weighted += volume * ctr
+
+        if total_volume > 0:
+            max_possible = total_volume * CTR_WEIGHTS[1]
+            return (total_weighted / max_possible) * 100.0
+        return 0.0
+
+    def generate_segment_report(self, trends: list[RankingTrend]) -> list[SegmentReport]:
+        """
+        Generate performance breakdown by keyword segment.
+
+        Segments include: brand, non_brand, and by intent type.
+        """
+        segment_map: dict[str, list[RankingTrend]] = {}
+
+        for trend in trends:
+            # Brand segment
+            brand_key = "brand" if trend.is_brand else "non_brand"
+            if brand_key not in segment_map:
+                segment_map[brand_key] = []
+            segment_map[brand_key].append(trend)
+
+            # Intent segment
+            intent_key = f"intent_{trend.intent.lower()}" if trend.intent else "intent_informational"
+            if intent_key not in segment_map:
+                segment_map[intent_key] = []
+            segment_map[intent_key].append(trend)
+
+        reports: list[SegmentReport] = []
+        for seg_name, seg_trends in sorted(segment_map.items()):
+            if not seg_trends:
+                continue
+
+            active = [t for t in seg_trends if t.current_position > 0]
+            avg_pos = sum(t.current_position for t in active) / len(active) if active else 0.0
+            avg_change = sum(t.total_change for t in seg_trends) / len(seg_trends) if seg_trends else 0.0
+
+            vis = self._calculate_visibility_score(seg_trends, use_start=False)
+
+            improved = sum(1 for t in seg_trends if t.trend_direction == "improved")
+            declined = sum(1 for t in seg_trends if t.trend_direction == "declined")
+            stable = sum(1 for t in seg_trends if t.trend_direction == "stable")
+
+            # Get top movers within segment
+            seg_gainers, seg_losers = self.find_top_movers(seg_trends, limit=5)
+
+            report = SegmentReport(
+                segment_name=seg_name,
+                total_keywords=len(seg_trends),
+                avg_position=round(avg_pos, 1),
+                avg_position_change=round(avg_change, 1),
+                visibility_score=round(vis, 2),
+                improved_count=improved,
+                declined_count=declined,
+                stable_count=stable,
+                top_gainers=seg_gainers,
+                top_losers=seg_losers,
+            )
+            reports.append(report)
+
+        return reports
+
+    async def compare_with_competitor(
+        self,
+        target: str,
+        competitor: str,
+        period_days: int = 30,
+    ) -> CompetitorReport:
+        """
+        Period-over-period comparison with a competitor.
+
+        Uses Ahrefs rank-tracker-competitors-stats for detailed comparison.
+        """
+        logger.info(f"Comparing {target} vs {competitor} over {period_days} days")
+
+        comp_data = await self._call_competitors_stats(target, competitor)
+
+        report = CompetitorReport(competitor=competitor)
+
+        if comp_data:
+            report.our_visibility = comp_data.get("target_visibility", 0.0)
+            report.their_visibility = comp_data.get("competitor_visibility", 0.0)
+            report.overlap_keywords = comp_data.get("overlap_keywords", 0)
+            report.keywords_we_lead = comp_data.get("target_better", 0)
+            report.keywords_they_lead = comp_data.get("competitor_better", 0)
+
+            # Extract notable gaps
+            gaps = comp_data.get("keyword_gaps", [])
+            report.notable_gaps = [
+                {
+                    "keyword": g.get("keyword", ""),
+                    "our_position": g.get("target_position", 0),
+                    "their_position": g.get("competitor_position", 0),
+                    "volume": g.get("volume", 0),
+                }
+                for g in gaps[:15]
+            ]
+
+        return report
+
+    async def _call_competitors_stats(self, target: str, competitor: str) -> dict:
+        """Call Ahrefs rank-tracker-competitors-stats MCP tool."""
+        logger.info("Calling Ahrefs rank-tracker-competitors-stats...")
+        try:
+            import subprocess
+            result = subprocess.run(
+                ["mcp-cli", "call", "ahrefs/rank-tracker-competitors-stats",
+                 json.dumps({"target": target, "competitor": competitor})],
+                capture_output=True, text=True, timeout=60,
+            )
+            if result.returncode == 0:
+                return json.loads(result.stdout)
+        except (FileNotFoundError, subprocess.TimeoutExpired, json.JSONDecodeError):
+            pass
+        return {}
+
+    async def generate_report(
+        self,
+        target: str,
+        period_days: int = 30,
+        competitors: Optional[list[str]] = None,
+    ) -> RankingReport:
+        """
+        Orchestrate full ranking performance report generation.
+
+        Args:
+            target: Target website URL
+            period_days: Reporting period in days
+            competitors: List of competitor URLs to compare
+
+        Returns:
+            Complete RankingReport with trends, movers, segments, and comparisons
+        """
+        logger.info(f"Generating ranking report for: {target} ({period_days} days)")
+
+        report = RankingReport(target=target, period_days=period_days)
+
+        # Step 1: Fetch historical position data
+        trends = await self.get_historical_positions(target, period_days)
+
+        if not trends:
+            logger.warning("No historical data retrieved. Check Ahrefs project configuration.")
+            return report
+
+        report.keyword_trends = trends
+        report.total_keywords = len(trends)
+
+        # Step 2: Calculate trend summary
+        report.trend_summary = self.calculate_trends(trends)
+
+        # Step 3: Calculate visibility scores (current vs period start)
+        report.current_visibility = self._calculate_visibility_score(trends, use_start=False)
+        report.previous_visibility = self._calculate_visibility_score(trends, use_start=True)
+        report.visibility_change = report.current_visibility - report.previous_visibility
+
+        # Step 4: Find top movers
+        gainers, losers = self.find_top_movers(trends, limit=10)
+        report.top_gainers = gainers
+        report.top_losers = losers
+
+        # Step 5: Generate segment reports
+        report.segments = self.generate_segment_report(trends)
+
+        # Step 6: Compare with competitors
+        if competitors:
+            for competitor in competitors:
+                comp_report = await self.compare_with_competitor(
+                    target, competitor, period_days,
+                )
+                report.competitors.append(comp_report)
+
+        logger.info(f"Report complete. Keywords: {report.total_keywords}")
+        logger.info(
+            f"Visibility: {report.previous_visibility:.2f} -> "
+            f"{report.current_visibility:.2f} ({report.visibility_change:+.2f})"
+        )
+
+        return report
+
+
+# ---------------------------------------------------------------------------
+# Output formatters
+# ---------------------------------------------------------------------------
+def format_text_report(report: RankingReport) -> str:
+    """Format ranking report as human-readable text."""
+    lines = []
+    lines.append("=" * 60)
+    lines.append(f"Ranking Performance Report: {report.target}")
+    lines.append(f"Period: {report.period_start} ~ {report.period_end} ({report.period_days} days)")
+    lines.append(f"Generated: {report.timestamp}")
+    lines.append("=" * 60)
+
+    # Visibility trend
+    lines.append(f"\nVisibility Score:")
+    lines.append(f"  Current:  {report.current_visibility:.2f}")
+    lines.append(f"  Previous: {report.previous_visibility:.2f}")
+    change_sign = "+" if report.visibility_change >= 0 else ""
+    lines.append(f"  Change:   {change_sign}{report.visibility_change:.2f}")
+
+    # Trend summary
+    ts = report.trend_summary
+    lines.append(f"\nKeyword Trends ({report.total_keywords} total):")
+    lines.append(f"  Improved: {ts.get('improved', 0)}")
+    lines.append(f"  Declined: {ts.get('declined', 0)}")
+    lines.append(f"  Stable:   {ts.get('stable', 0)}")
+    lines.append(f"  New:      {ts.get('new', 0)}")
+    lines.append(f"  Lost:     {ts.get('lost', 0)}")
+
+    # Top gainers
+    if report.top_gainers:
+        lines.append(f"\nTop Gainers:")
+        lines.append("-" * 60)
+        for m in report.top_gainers:
+            lines.append(
+                f"  {m.keyword}: {m.previous_position} -> {m.current_position} "
+                f"(+{m.position_change}) | Vol: {m.volume} | Impact: {m.impact_score}"
+            )
+
+    # Top losers
+    if report.top_losers:
+        lines.append(f"\nTop Losers:")
+        lines.append("-" * 60)
+        for m in report.top_losers:
+            lines.append(
+                f"  {m.keyword}: {m.previous_position} -> {m.current_position} "
+                f"({m.position_change}) | Vol: {m.volume} | Impact: {m.impact_score}"
+            )
+
+    # Segments
+    if report.segments:
+        lines.append(f"\nSegment Breakdown:")
+        lines.append("-" * 60)
+        for seg in report.segments:
+            lines.append(
+                f"  {seg.segment_name}: {seg.total_keywords} kw, "
+                f"avg pos {seg.avg_position}, vis {seg.visibility_score}, "
+                f"improved {seg.improved_count} / declined {seg.declined_count}"
+            )
+
+    # Competitors
+    if report.competitors:
+        lines.append(f"\nCompetitor Comparison:")
+        lines.append("-" * 60)
+        for comp in report.competitors:
+            lines.append(f"  vs {comp.competitor}:")
+            lines.append(f"    Our visibility:   {comp.our_visibility:.2f}")
+            lines.append(f"    Their visibility: {comp.their_visibility:.2f}")
+            lines.append(f"    Overlap: {comp.overlap_keywords} keywords")
+            lines.append(f"    We lead: {comp.keywords_we_lead}")
+            lines.append(f"    They lead: {comp.keywords_they_lead}")
+            if comp.notable_gaps:
+                lines.append(f"    Notable gaps:")
+                for gap in comp.notable_gaps[:5]:
+                    lines.append(
+                        f"      {gap['keyword']}: us #{gap['our_position']} "
+                        f"vs them #{gap['their_position']} (vol: {gap['volume']})"
+                    )
+
+    lines.append("\n" + "=" * 60)
+    return "\n".join(lines)
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Ranking Reporter - Generate ranking performance reports with trends",
+    )
+    parser.add_argument(
+        "--target",
+        required=True,
+        help="Target website URL (e.g., https://example.com)",
+    )
+    parser.add_argument(
+        "--period",
+        type=int,
+        default=30,
+        help="Reporting period in days (default: 30)",
+    )
+    parser.add_argument(
+        "--competitor",
+        action="append",
+        dest="competitors",
+        default=[],
+        help="Competitor URL to compare (repeatable)",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        dest="json_output",
+        help="Output in JSON format",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        default=None,
+        help="Save output to file path",
+    )
+    return parser.parse_args()
+
+
+async def main():
+    args = parse_args()
+
+    reporter = RankingReporter()
+
+    report = await reporter.generate_report(
+        target=args.target,
+        period_days=args.period,
+        competitors=args.competitors,
+    )
+
+    if args.json_output:
+        output = json.dumps(report.to_dict(), ensure_ascii=False, indent=2)
+    else:
+        output = format_text_report(report)
+
+    if args.output:
+        with open(args.output, "w", encoding="utf-8") as f:
+            f.write(output)
+        logger.info(f"Output saved to: {args.output}")
+    else:
+        print(output)
+
+    reporter.print_stats()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/21-seo-position-tracking/code/scripts/requirements.txt b/custom-skills/21-seo-position-tracking/code/scripts/requirements.txt
new file mode 100644
index 0000000..c355246
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/code/scripts/requirements.txt
@@ -0,0 +1,8 @@
+# 21-seo-position-tracking dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+pandas>=2.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/21-seo-position-tracking/desktop/SKILL.md b/custom-skills/21-seo-position-tracking/desktop/SKILL.md
new file mode 100644
index 0000000..4e109cc
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/desktop/SKILL.md
@@ -0,0 +1,107 @@
+---
+name: seo-position-tracking
+description: |
+  Keyword position tracking and ranking monitoring via Ahrefs Rank Tracker.
+  Triggers: rank tracking, position monitoring, keyword rankings, visibility score, ranking report, 키워드 순위, 순위 추적.
+---
+
+# SEO Position Tracking
+
+## Purpose
+
+Monitor keyword ranking positions, detect significant changes, calculate visibility scores, and compare against competitors using Ahrefs Rank Tracker data. Provides actionable alerts for ranking drops and segment-level performance breakdown.
+
+## Core Capabilities
+
+1. **Position Monitoring** - Retrieve current keyword ranking positions from Ahrefs Rank Tracker projects
+2. **Change Detection** - Detect significant position changes with configurable threshold alerts (severity: critical/high/medium/low)
+3. **Visibility Scoring** - Calculate weighted visibility scores using CTR-curve model (position 1 = 30%, position 2 = 15%, etc.)
+4. **Brand/Non-brand Segmentation** - Automatically classify keywords by brand relevance and search intent type
+5. **Competitor Comparison** - Compare keyword overlap, position gaps, and visibility scores against competitors
+
+## MCP Tool Usage
+
+### Ahrefs Rank Tracker Tools
+```
+mcp__ahrefs__rank-tracker-overview: Get rank tracking overview with current positions
+mcp__ahrefs__rank-tracker-competitors-overview: Compare rankings against competitors
+mcp__ahrefs__rank-tracker-competitors-pages: Competitor page-level ranking data
+mcp__ahrefs__rank-tracker-competitors-stats: Detailed competitor ranking statistics
+mcp__ahrefs__rank-tracker-serp-overview: SERP details for tracked keywords
+mcp__ahrefs__management-projects: List available Ahrefs projects
+mcp__ahrefs__management-project-keywords: Get tracked keywords for a project
+```
+
+### Notion for Report Storage
+```
+mcp__notion__notion-create-pages: Save tracking reports to SEO Audit Log
+mcp__notion__notion-update-page: Update existing tracking entries
+```
+
+## Workflow
+
+### Phase 1: Data Collection
+1. Identify Ahrefs project via `management-projects`
+2. Retrieve tracked keywords via `management-project-keywords`
+3. Fetch current positions via `rank-tracker-overview`
+4. Fetch competitor data via `rank-tracker-competitors-overview` (if requested)
+
+### Phase 2: Analysis
+1. Detect position changes against previous period
+2. Generate alerts for changes exceeding threshold
+3. Calculate visibility score weighted by search volume and CTR curve
+4. Segment keywords into brand/non-brand and by intent type
+5. Compare positions against each competitor
+
+### Phase 3: Reporting
+1. Compile position distribution (top3/top10/top20/top50/top100)
+2. Summarize changes (improved/declined/stable/new/lost)
+3. List alerts sorted by severity and search volume
+4. Generate segment-level breakdown
+5. Save report to Notion SEO Audit Log database
+
+## Output Format
+
+```json
+{
+  "target": "https://example.com",
+  "total_keywords": 250,
+  "visibility_score": 68.5,
+  "positions": {
+    "top3": 15,
+    "top10": 48,
+    "top20": 92,
+    "top50": 180,
+    "top100": 230
+  },
+  "changes": {
+    "improved": 45,
+    "declined": 30,
+    "stable": 155,
+    "new": 12,
+    "lost": 8
+  },
+  "alerts": [
+    {
+      "keyword": "example keyword",
+      "old_position": 5,
+      "new_position": 15,
+      "change": -10,
+      "volume": 5400,
+      "severity": "high"
+    }
+  ],
+  "segments": {
+    "brand": {"keywords": 30, "avg_position": 2.1},
+    "non_brand": {"keywords": 220, "avg_position": 24.5}
+  }
+}
+```
+
+## Notion Output (Required)
+
+All tracking reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category (Position Tracking), Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: RANK-YYYYMMDD-NNN
diff --git a/custom-skills/21-seo-position-tracking/desktop/skill.yaml b/custom-skills/21-seo-position-tracking/desktop/skill.yaml
new file mode 100644
index 0000000..37cd9a2
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-position-tracking
+description: |
+  Keyword position tracking and ranking monitoring. Triggers: rank tracking, position monitoring, keyword rankings, visibility score, ranking report.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/21-seo-position-tracking/desktop/tools/ahrefs.md b/custom-skills/21-seo-position-tracking/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..5d918e0
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/desktop/tools/ahrefs.md
@@ -0,0 +1,15 @@
+# Ahrefs
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/21-seo-position-tracking/desktop/tools/notion.md b/custom-skills/21-seo-position-tracking/desktop/tools/notion.md
new file mode 100644
index 0000000..153e086
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/desktop/tools/notion.md
@@ -0,0 +1,15 @@
+# Notion
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/21-seo-position-tracking/desktop/tools/websearch.md b/custom-skills/21-seo-position-tracking/desktop/tools/websearch.md
new file mode 100644
index 0000000..36ef238
--- /dev/null
+++ b/custom-skills/21-seo-position-tracking/desktop/tools/websearch.md
@@ -0,0 +1,15 @@
+# WebSearch
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/22-seo-link-building/code/CLAUDE.md b/custom-skills/22-seo-link-building/code/CLAUDE.md
new file mode 100644
index 0000000..24498c7
--- /dev/null
+++ b/custom-skills/22-seo-link-building/code/CLAUDE.md
@@ -0,0 +1,144 @@
+# CLAUDE.md
+
+## Overview
+
+Link building diagnosis tool for backlink profile analysis, toxic link detection, competitor link gap identification, and link velocity tracking. Supports Korean platform link mapping (Naver Blog, Cafe, Tistory, Brunch, Korean news sites).
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Backlink profile audit
+python scripts/backlink_auditor.py --url https://example.com --json
+
+# Link gap analysis vs competitors
+python scripts/link_gap_finder.py --target https://example.com --competitor https://competitor.com --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `backlink_auditor.py` | Analyze backlink profile, detect toxic links | DR, referring domains, anchor distribution, toxic links |
+| `link_gap_finder.py` | Find link gap opportunities vs competitors | Domains linking to competitors but not target |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## Backlink Auditor
+
+```bash
+# Full backlink audit
+python scripts/backlink_auditor.py --url https://example.com --json
+
+# Check link velocity (new/lost over time)
+python scripts/backlink_auditor.py --url https://example.com --velocity --json
+
+# Find broken backlinks for recovery
+python scripts/backlink_auditor.py --url https://example.com --broken --json
+
+# Korean platform link analysis
+python scripts/backlink_auditor.py --url https://example.com --korean-platforms --json
+```
+
+**Capabilities**:
+- Domain Rating (DR) and backlink stats overview
+- Referring domain analysis (count, DR distribution, country distribution)
+- Anchor text distribution analysis (branded, exact-match, generic, naked URL)
+- Toxic link detection (PBN patterns, spammy domains, link farms)
+- Link velocity tracking (new/lost referring domains over time)
+- Broken backlink recovery opportunities
+- Korean platform mapping (Naver Blog, Naver Cafe, Tistory, Brunch, Korean news)
+
+## Link Gap Finder
+
+```bash
+# Gap vs one competitor
+python scripts/link_gap_finder.py --target https://example.com --competitor https://comp1.com --json
+
+# Multiple competitors
+python scripts/link_gap_finder.py --target https://example.com --competitor https://comp1.com --competitor https://comp2.com --json
+
+# Filter by minimum DR
+python scripts/link_gap_finder.py --target https://example.com --competitor https://comp1.com --min-dr 30 --json
+```
+
+**Capabilities**:
+- Find domains linking to competitors but not target
+- Score link opportunities by DR, traffic, relevance
+- Categorize link sources (editorial, directory, forum, blog, news)
+- Prioritize by feasibility and impact
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `site-explorer-all-backlinks` | Get all backlinks for a target |
+| `site-explorer-backlinks-stats` | Backlink statistics overview |
+| `site-explorer-referring-domains` | List referring domains |
+| `site-explorer-anchors` | Anchor text distribution |
+| `site-explorer-broken-backlinks` | Find broken backlinks |
+| `site-explorer-domain-rating` | Get Domain Rating |
+| `site-explorer-domain-rating-history` | DR trend over time |
+| `site-explorer-refdomains-history` | Referring domains trend |
+| `site-explorer-linked-domains` | Domains linked from target |
+
+## Output Format
+
+```json
+{
+  "url": "https://example.com",
+  "domain_rating": 45,
+  "backlink_stats": {
+    "total_backlinks": 12500,
+    "referring_domains": 850,
+    "dofollow_ratio": 0.72
+  },
+  "anchor_distribution": {
+    "branded": 35,
+    "exact_match": 12,
+    "partial_match": 18,
+    "generic": 20,
+    "naked_url": 15
+  },
+  "toxic_links": [...],
+  "korean_platforms": {
+    "naver_blog": 45,
+    "naver_cafe": 12,
+    "tistory": 23,
+    "brunch": 5
+  },
+  "link_velocity": {
+    "new_last_30d": 120,
+    "lost_last_30d": 35
+  },
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Audited website URL |
+| Category | Select | Link Building |
+| Priority | Select | Based on toxic link count and gap size |
+| Found Date | Date | Audit date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: LINK-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., Domain Rating, Referring Domains, Backlinks)
+- URLs and code remain unchanged
diff --git a/custom-skills/22-seo-link-building/code/scripts/backlink_auditor.py b/custom-skills/22-seo-link-building/code/scripts/backlink_auditor.py
new file mode 100644
index 0000000..aa97616
--- /dev/null
+++ b/custom-skills/22-seo-link-building/code/scripts/backlink_auditor.py
@@ -0,0 +1,1079 @@
+"""
+Backlink Auditor - Backlink profile analysis and toxic link detection
+=====================================================================
+Purpose: Analyze backlink profiles via Ahrefs MCP, detect toxic links,
+         track link velocity, and map Korean platform links.
+Python: 3.10+
+Usage:
+    python backlink_auditor.py --url https://example.com --json
+    python backlink_auditor.py --url https://example.com --velocity --broken --json
+    python backlink_auditor.py --url https://example.com --korean-platforms --json
+"""
+
+from __future__ import annotations
+
+import argparse
+import asyncio
+import json
+import logging
+import re
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime, timedelta
+from typing import Any
+from urllib.parse import urlparse
+
+import aiohttp
+import pandas as pd
+from rich.console import Console
+from rich.table import Table
+
+from base_client import BaseAsyncClient, config
+
+# ---------------------------------------------------------------------------
+# Logging
+# ---------------------------------------------------------------------------
+logger = logging.getLogger("backlink_auditor")
+console = Console()
+
+# ---------------------------------------------------------------------------
+# Constants
+# ---------------------------------------------------------------------------
+AHREFS_BASE = "https://api.ahrefs.com/v3"
+
+# Korean platform domain patterns
+KOREAN_PLATFORM_PATTERNS: dict[str, list[str]] = {
+    "naver_blog": ["blog.naver.com"],
+    "naver_cafe": ["cafe.naver.com"],
+    "naver_post": ["post.naver.com"],
+    "naver_kin": ["kin.naver.com"],
+    "tistory": ["tistory.com"],
+    "brunch": ["brunch.co.kr"],
+    "daum_blog": ["blog.daum.net"],
+    "korean_news": [
+        "chosun.com", "donga.com", "joongang.co.kr", "hani.co.kr",
+        "khan.co.kr", "mk.co.kr", "mt.co.kr", "hankyung.com",
+        "sedaily.com", "edaily.co.kr", "newsis.com", "yna.co.kr",
+        "yonhapnews.co.kr", "news1.kr", "newspim.com", "etnews.com",
+        "zdnet.co.kr", "bloter.net", "platum.kr", "besuccess.com",
+    ],
+    "korean_community": [
+        "dcinside.com", "theqoo.net", "fmkorea.com", "ruliweb.com",
+        "ppomppu.co.kr", "clien.net", "mlbpark.donga.com",
+    ],
+}
+
+# Suspicious TLD patterns for toxic link detection
+SUSPICIOUS_TLDS = {
+    ".xyz", ".top", ".club", ".work", ".date", ".bid",
+    ".stream", ".download", ".win", ".racing", ".review",
+    ".accountant", ".loan", ".click", ".link", ".gdn",
+}
+
+# Generic anchor text patterns
+GENERIC_ANCHORS = {
+    "click here", "read more", "learn more", "visit", "here",
+    "website", "source", "link", "this", "more info",
+    "go to", "check out", "see more", "view", "details",
+}
+
+# ---------------------------------------------------------------------------
+# Dataclasses
+# ---------------------------------------------------------------------------
+
+
+@dataclass
+class BacklinkEntry:
+    """Single backlink record."""
+    source_url: str
+    source_domain: str
+    dr: float = 0.0
+    anchor: str = ""
+    link_type: str = "text"
+    dofollow: bool = True
+    first_seen: str = ""
+    url_to: str = ""
+
+
+@dataclass
+class AnchorDistribution:
+    """Anchor text distribution breakdown."""
+    branded: int = 0
+    exact_match: int = 0
+    partial_match: int = 0
+    generic: int = 0
+    naked_url: int = 0
+    other: int = 0
+    total: int = 0
+    top_anchors: list[dict[str, Any]] = field(default_factory=list)
+
+    @property
+    def branded_pct(self) -> float:
+        return round(self.branded / max(self.total, 1) * 100, 1)
+
+    @property
+    def exact_match_pct(self) -> float:
+        return round(self.exact_match / max(self.total, 1) * 100, 1)
+
+    @property
+    def generic_pct(self) -> float:
+        return round(self.generic / max(self.total, 1) * 100, 1)
+
+    @property
+    def naked_url_pct(self) -> float:
+        return round(self.naked_url / max(self.total, 1) * 100, 1)
+
+
+@dataclass
+class ToxicLink:
+    """Potentially toxic or spammy backlink."""
+    url: str
+    domain: str
+    reason: str
+    risk_score: float = 0.0
+    dr: float = 0.0
+    anchor: str = ""
+    first_seen: str = ""
+
+
+@dataclass
+class KoreanPlatformStats:
+    """Backlink counts from Korean platforms."""
+    naver_blog: int = 0
+    naver_cafe: int = 0
+    naver_post: int = 0
+    naver_kin: int = 0
+    tistory: int = 0
+    brunch: int = 0
+    daum_blog: int = 0
+    korean_news: int = 0
+    korean_community: int = 0
+    total_korean: int = 0
+    korean_domains: list[dict[str, Any]] = field(default_factory=list)
+
+
+@dataclass
+class LinkVelocity:
+    """Link acquisition and loss velocity."""
+    new_last_7d: int = 0
+    new_last_30d: int = 0
+    new_last_90d: int = 0
+    lost_last_7d: int = 0
+    lost_last_30d: int = 0
+    lost_last_90d: int = 0
+    velocity_trend: str = "stable"
+    history: list[dict[str, Any]] = field(default_factory=list)
+
+
+@dataclass
+class BacklinkAuditResult:
+    """Complete backlink audit result."""
+    url: str
+    domain: str = ""
+    domain_rating: float = 0.0
+    total_backlinks: int = 0
+    referring_domains: int = 0
+    dofollow_backlinks: int = 0
+    nofollow_backlinks: int = 0
+    dofollow_ratio: float = 0.0
+    edu_gov_backlinks: int = 0
+    anchor_distribution: AnchorDistribution | None = None
+    toxic_links: list[ToxicLink] = field(default_factory=list)
+    toxic_link_count: int = 0
+    toxic_risk_level: str = "low"
+    broken_backlinks: list[dict[str, Any]] = field(default_factory=list)
+    korean_platforms: KoreanPlatformStats | None = None
+    link_velocity: LinkVelocity | None = None
+    dr_distribution: dict[str, int] = field(default_factory=dict)
+    country_distribution: list[dict[str, Any]] = field(default_factory=list)
+    issues: list[dict[str, str]] = field(default_factory=list)
+    recommendations: list[str] = field(default_factory=list)
+    timestamp: str = ""
+
+
+# ---------------------------------------------------------------------------
+# BacklinkAuditor
+# ---------------------------------------------------------------------------
+
+
+class BacklinkAuditor(BaseAsyncClient):
+    """Analyze backlink profiles using Ahrefs MCP tools."""
+
+    def __init__(self, **kwargs):
+        super().__init__(max_concurrent=5, requests_per_second=2.0, **kwargs)
+        self.session: aiohttp.ClientSession | None = None
+
+    # -- Ahrefs MCP helper ---------------------------------------------------
+
+    async def _call_ahrefs(
+        self, endpoint: str, params: dict[str, Any]
+    ) -> dict[str, Any]:
+        """
+        Call Ahrefs API endpoint.
+
+        In MCP context this would call mcp__ahrefs__<endpoint>.
+        For standalone use, falls back to REST API with token.
+        """
+        api_token = config.get_required("AHREFS_API_TOKEN") if not self.session else None
+
+        if self.session and api_token:
+            url = f"{AHREFS_BASE}/{endpoint}"
+            headers = {"Authorization": f"Bearer {api_token}"}
+            async with self.session.get(url, headers=headers, params=params) as resp:
+                resp.raise_for_status()
+                return await resp.json()
+
+        # Fallback: return structured empty result for MCP-only usage
+        logger.warning(
+            f"Ahrefs call to '{endpoint}' - use MCP tool "
+            f"mcp__ahrefs__{endpoint.replace('-', '_')} in Claude Desktop"
+        )
+        return {"endpoint": endpoint, "params": params, "data": [], "note": "mcp_stub"}
+
+    # -- Core methods --------------------------------------------------------
+
+    async def get_backlink_stats(self, url: str) -> dict[str, Any]:
+        """Fetch backlink statistics overview via site-explorer-backlinks-stats."""
+        target = urlparse(url).netloc or url
+        result = await self._call_ahrefs(
+            "site-explorer-backlinks-stats",
+            {"target": target, "mode": "domain"},
+        )
+        stats = result.get("data", result) if isinstance(result, dict) else {}
+        return {
+            "total_backlinks": stats.get("live", 0),
+            "referring_domains": stats.get("live_refdomains", 0),
+            "dofollow": stats.get("live_dofollow", 0),
+            "nofollow": stats.get("live_nofollow", 0),
+            "edu": stats.get("edu", 0),
+            "gov": stats.get("gov", 0),
+            "raw": stats,
+        }
+
+    async def get_domain_rating(self, url: str) -> dict[str, Any]:
+        """Fetch Domain Rating via site-explorer-domain-rating."""
+        target = urlparse(url).netloc or url
+        result = await self._call_ahrefs(
+            "site-explorer-domain-rating",
+            {"target": target},
+        )
+        data = result.get("data", result) if isinstance(result, dict) else {}
+        return {
+            "domain_rating": data.get("domain_rating", 0),
+            "ahrefs_rank": data.get("ahrefs_rank", 0),
+        }
+
+    async def get_referring_domains(
+        self, url: str, limit: int = 1000
+    ) -> list[dict[str, Any]]:
+        """List referring domains via site-explorer-referring-domains."""
+        target = urlparse(url).netloc or url
+        result = await self._call_ahrefs(
+            "site-explorer-referring-domains",
+            {"target": target, "mode": "domain", "limit": limit, "order_by": "domain_rating:desc"},
+        )
+        domains = result.get("data", result.get("refdomains", []))
+        if isinstance(domains, dict):
+            domains = domains.get("refdomains", [])
+        return domains if isinstance(domains, list) else []
+
+    async def get_all_backlinks(
+        self, url: str, limit: int = 1000
+    ) -> list[BacklinkEntry]:
+        """Fetch all backlinks via site-explorer-all-backlinks."""
+        target = urlparse(url).netloc or url
+        result = await self._call_ahrefs(
+            "site-explorer-all-backlinks",
+            {"target": target, "mode": "domain", "limit": limit, "order_by": "domain_rating:desc"},
+        )
+        raw_links = result.get("data", result.get("backlinks", []))
+        if isinstance(raw_links, dict):
+            raw_links = raw_links.get("backlinks", [])
+
+        backlinks = []
+        for link in (raw_links if isinstance(raw_links, list) else []):
+            entry = BacklinkEntry(
+                source_url=link.get("url_from", ""),
+                source_domain=link.get("domain_from", ""),
+                dr=link.get("domain_rating", 0),
+                anchor=link.get("anchor", ""),
+                link_type=link.get("type", "text"),
+                dofollow=not link.get("nofollow", False),
+                first_seen=link.get("first_seen", ""),
+                url_to=link.get("url_to", ""),
+            )
+            backlinks.append(entry)
+        return backlinks
+
+    async def analyze_anchors(
+        self, url: str, brand_name: str = ""
+    ) -> AnchorDistribution:
+        """Analyze anchor text distribution via site-explorer-anchors."""
+        target = urlparse(url).netloc or url
+        if not brand_name:
+            brand_name = target.replace("www.", "").split(".")[0]
+
+        result = await self._call_ahrefs(
+            "site-explorer-anchors",
+            {"target": target, "mode": "domain", "limit": 500, "order_by": "backlinks:desc"},
+        )
+        raw_anchors = result.get("data", result.get("anchors", []))
+        if isinstance(raw_anchors, dict):
+            raw_anchors = raw_anchors.get("anchors", [])
+
+        dist = AnchorDistribution()
+        top_list = []
+
+        for item in (raw_anchors if isinstance(raw_anchors, list) else []):
+            anchor_text = item.get("anchor", "").strip().lower()
+            count = item.get("backlinks", 1)
+            dist.total += count
+
+            top_list.append({
+                "anchor": item.get("anchor", ""),
+                "backlinks": count,
+                "referring_domains": item.get("refdomains", 0),
+            })
+
+            # Classify anchor text
+            if self._is_branded_anchor(anchor_text, brand_name, target):
+                dist.branded += count
+            elif self._is_naked_url(anchor_text, target):
+                dist.naked_url += count
+            elif self._is_generic_anchor(anchor_text):
+                dist.generic += count
+            elif self._is_exact_match(anchor_text, brand_name):
+                dist.exact_match += count
+            else:
+                dist.partial_match += count
+
+        dist.top_anchors = sorted(top_list, key=lambda x: x["backlinks"], reverse=True)[:20]
+        return dist
+
+    async def detect_toxic_links(
+        self, backlinks: list[BacklinkEntry]
+    ) -> list[ToxicLink]:
+        """Identify potentially toxic backlinks using heuristic scoring."""
+        toxic_links = []
+
+        for link in backlinks:
+            risk_score = 0.0
+            reasons = []
+
+            # Check 1: Suspicious TLD
+            domain_tld = self._extract_tld(link.source_domain)
+            if domain_tld in SUSPICIOUS_TLDS:
+                risk_score += 25.0
+                reasons.append(f"Suspicious TLD: {domain_tld}")
+
+            # Check 2: Very low DR (potential PBN or link farm)
+            if link.dr < 5 and link.source_domain:
+                risk_score += 20.0
+                reasons.append(f"Very low DR ({link.dr})")
+            elif link.dr < 10:
+                risk_score += 10.0
+                reasons.append(f"Low DR ({link.dr})")
+
+            # Check 3: Spammy anchor patterns
+            anchor_lower = link.anchor.lower().strip()
+            spam_keywords = [
+                "casino", "poker", "viagra", "cialis", "payday",
+                "loan", "buy cheap", "free download", "adult",
+                "gambling", "betting", "porn", "xxx", "sex",
+                "weight loss", "diet pill", "crypto trading",
+            ]
+            for kw in spam_keywords:
+                if kw in anchor_lower:
+                    risk_score += 30.0
+                    reasons.append(f"Spam keyword in anchor: '{kw}'")
+                    break
+
+            # Check 4: Numeric/random domain patterns (e.g., abc123xyz.com)
+            domain_base = link.source_domain.split(".")[0] if link.source_domain else ""
+            if domain_base and re.match(r"^[a-z0-9]{15,}$", domain_base):
+                risk_score += 15.0
+                reasons.append("Random/generated domain name")
+
+            # Check 5: Excessive hyphens in domain
+            if domain_base.count("-") >= 3:
+                risk_score += 15.0
+                reasons.append("Excessive hyphens in domain")
+
+            # Check 6: Domain contains spam-related words
+            domain_spam_words = [
+                "seo", "link", "backlink", "directory", "submit",
+                "free-", "cheap-", "best-", "buy-",
+            ]
+            for sw in domain_spam_words:
+                if sw in link.source_domain.lower():
+                    risk_score += 10.0
+                    reasons.append(f"Spam word in domain: '{sw}'")
+                    break
+
+            # Check 7: Very long domain name
+            if len(domain_base) > 30:
+                risk_score += 10.0
+                reasons.append("Unusually long domain name")
+
+            # Threshold: toxic if score >= 30
+            if risk_score >= 30.0:
+                toxic = ToxicLink(
+                    url=link.source_url,
+                    domain=link.source_domain,
+                    reason="; ".join(reasons),
+                    risk_score=min(risk_score, 100.0),
+                    dr=link.dr,
+                    anchor=link.anchor,
+                    first_seen=link.first_seen,
+                )
+                toxic_links.append(toxic)
+
+        # Sort by risk score descending
+        toxic_links.sort(key=lambda t: t.risk_score, reverse=True)
+        return toxic_links
+
+    async def find_broken_backlinks(self, url: str) -> list[dict[str, Any]]:
+        """Find broken backlinks for recovery via site-explorer-broken-backlinks."""
+        target = urlparse(url).netloc or url
+        result = await self._call_ahrefs(
+            "site-explorer-broken-backlinks",
+            {"target": target, "mode": "domain", "limit": 200, "order_by": "domain_rating:desc"},
+        )
+        raw = result.get("data", result.get("backlinks", []))
+        if isinstance(raw, dict):
+            raw = raw.get("backlinks", [])
+
+        broken = []
+        for item in (raw if isinstance(raw, list) else []):
+            broken.append({
+                "source_url": item.get("url_from", ""),
+                "source_domain": item.get("domain_from", ""),
+                "target_url": item.get("url_to", ""),
+                "http_code": item.get("http_code", 404),
+                "anchor": item.get("anchor", ""),
+                "dr": item.get("domain_rating", 0),
+                "first_seen": item.get("first_seen", ""),
+            })
+
+        broken.sort(key=lambda b: b.get("dr", 0), reverse=True)
+        return broken
+
+    async def track_velocity(self, url: str) -> LinkVelocity:
+        """Track new/lost referring domains via refdomains-history."""
+        target = urlparse(url).netloc or url
+        today = datetime.now()
+
+        result = await self._call_ahrefs(
+            "site-explorer-refdomains-history",
+            {
+                "target": target,
+                "mode": "domain",
+                "date_from": (today - timedelta(days=90)).strftime("%Y-%m-%d"),
+            },
+        )
+        raw_history = result.get("data", result.get("history", []))
+        if isinstance(raw_history, dict):
+            raw_history = raw_history.get("history", [])
+
+        velocity = LinkVelocity()
+        history_points = []
+
+        if isinstance(raw_history, list) and len(raw_history) >= 2:
+            for point in raw_history:
+                history_points.append({
+                    "date": point.get("date", ""),
+                    "referring_domains": point.get("refdomains", point.get("referring_domains", 0)),
+                })
+
+            # Calculate velocity from history deltas
+            sorted_history = sorted(history_points, key=lambda h: h["date"])
+
+            if len(sorted_history) >= 2:
+                latest = sorted_history[-1].get("referring_domains", 0)
+                d7_ago = self._find_closest_point(sorted_history, 7)
+                d30_ago = self._find_closest_point(sorted_history, 30)
+                d90_ago = self._find_closest_point(sorted_history, 90)
+
+                velocity.new_last_7d = max(0, latest - d7_ago)
+                velocity.new_last_30d = max(0, latest - d30_ago)
+                velocity.new_last_90d = max(0, latest - d90_ago)
+
+                # Estimate lost (simplified: if delta is negative)
+                velocity.lost_last_7d = max(0, d7_ago - latest)
+                velocity.lost_last_30d = max(0, d30_ago - latest)
+                velocity.lost_last_90d = max(0, d90_ago - latest)
+
+                # Determine trend
+                if velocity.new_last_30d > velocity.lost_last_30d * 2:
+                    velocity.velocity_trend = "growing"
+                elif velocity.lost_last_30d > velocity.new_last_30d * 2:
+                    velocity.velocity_trend = "declining"
+                else:
+                    velocity.velocity_trend = "stable"
+
+        velocity.history = history_points[-30:]  # Last 30 data points
+        return velocity
+
+    def map_korean_platforms(
+        self, referring_domains: list[dict[str, Any]]
+    ) -> KoreanPlatformStats:
+        """Categorize referring domains by Korean platform."""
+        stats = KoreanPlatformStats()
+        korean_detail = []
+
+        for rd in referring_domains:
+            domain = rd.get("domain", rd.get("domain_from", "")).lower()
+            matched_platform = None
+
+            for platform, patterns in KOREAN_PLATFORM_PATTERNS.items():
+                for pattern in patterns:
+                    if pattern in domain:
+                        matched_platform = platform
+                        break
+                if matched_platform:
+                    break
+
+            if matched_platform:
+                current_val = getattr(stats, matched_platform, 0)
+                setattr(stats, matched_platform, current_val + 1)
+                stats.total_korean += 1
+                korean_detail.append({
+                    "domain": domain,
+                    "platform": matched_platform,
+                    "dr": rd.get("domain_rating", rd.get("dr", 0)),
+                    "backlinks": rd.get("backlinks", 0),
+                })
+
+        stats.korean_domains = sorted(
+            korean_detail, key=lambda d: d.get("dr", 0), reverse=True
+        )
+        return stats
+
+    # -- Orchestration -------------------------------------------------------
+
+    async def audit(
+        self,
+        url: str,
+        include_velocity: bool = False,
+        include_broken: bool = False,
+        include_korean: bool = False,
+        brand_name: str = "",
+    ) -> BacklinkAuditResult:
+        """Orchestrate a full backlink audit."""
+        domain = urlparse(url).netloc or url
+        logger.info(f"Starting backlink audit for: {domain}")
+
+        result = BacklinkAuditResult(
+            url=url,
+            domain=domain,
+            timestamp=datetime.now().isoformat(),
+        )
+
+        # Phase 1: Core metrics (parallel)
+        logger.info("Phase 1: Fetching core metrics...")
+        stats_task = self.get_backlink_stats(url)
+        dr_task = self.get_domain_rating(url)
+        anchors_task = self.analyze_anchors(url, brand_name)
+        backlinks_task = self.get_all_backlinks(url, limit=1000)
+        refdomains_task = self.get_referring_domains(url, limit=1000)
+
+        stats, dr_info, anchors, backlinks, refdomains = await asyncio.gather(
+            stats_task, dr_task, anchors_task, backlinks_task, refdomains_task,
+            return_exceptions=True,
+        )
+
+        # Process stats
+        if isinstance(stats, dict):
+            result.total_backlinks = stats.get("total_backlinks", 0)
+            result.referring_domains = stats.get("referring_domains", 0)
+            result.dofollow_backlinks = stats.get("dofollow", 0)
+            result.nofollow_backlinks = stats.get("nofollow", 0)
+            result.edu_gov_backlinks = stats.get("edu", 0) + stats.get("gov", 0)
+            total = result.dofollow_backlinks + result.nofollow_backlinks
+            result.dofollow_ratio = round(
+                result.dofollow_backlinks / max(total, 1), 2
+            )
+
+        # Process DR
+        if isinstance(dr_info, dict):
+            result.domain_rating = dr_info.get("domain_rating", 0)
+
+        # Process anchors
+        if isinstance(anchors, AnchorDistribution):
+            result.anchor_distribution = anchors
+
+        # Phase 2: Toxic link detection
+        if isinstance(backlinks, list):
+            logger.info("Phase 2: Detecting toxic links...")
+            result.toxic_links = await self.detect_toxic_links(backlinks)
+            result.toxic_link_count = len(result.toxic_links)
+
+            # Determine toxic risk level
+            if result.toxic_link_count > 50:
+                result.toxic_risk_level = "critical"
+            elif result.toxic_link_count > 20:
+                result.toxic_risk_level = "high"
+            elif result.toxic_link_count > 5:
+                result.toxic_risk_level = "medium"
+            else:
+                result.toxic_risk_level = "low"
+
+        # Phase 3: DR distribution of referring domains
+        if isinstance(refdomains, list):
+            dr_dist = {"0-10": 0, "11-20": 0, "21-30": 0, "31-40": 0,
+                       "41-50": 0, "51-60": 0, "61-70": 0, "71-80": 0,
+                       "81-90": 0, "91-100": 0}
+            country_counts: dict[str, int] = {}
+
+            for rd in refdomains:
+                dr_val = rd.get("domain_rating", rd.get("dr", 0))
+                bucket = self._dr_bucket(dr_val)
+                dr_dist[bucket] = dr_dist.get(bucket, 0) + 1
+
+                country = rd.get("country", "unknown")
+                country_counts[country] = country_counts.get(country, 0) + 1
+
+            result.dr_distribution = dr_dist
+            result.country_distribution = sorted(
+                [{"country": k, "count": v} for k, v in country_counts.items()],
+                key=lambda x: x["count"], reverse=True,
+            )[:20]
+
+        # Phase 4: Optional analyses
+        if include_velocity:
+            logger.info("Phase 4a: Tracking link velocity...")
+            velocity = await self.track_velocity(url)
+            if isinstance(velocity, LinkVelocity):
+                result.link_velocity = velocity
+
+        if include_broken:
+            logger.info("Phase 4b: Finding broken backlinks...")
+            broken = await self.find_broken_backlinks(url)
+            if isinstance(broken, list):
+                result.broken_backlinks = broken
+
+        if include_korean and isinstance(refdomains, list):
+            logger.info("Phase 4c: Mapping Korean platforms...")
+            korean_stats = self.map_korean_platforms(refdomains)
+            result.korean_platforms = korean_stats
+
+        # Phase 5: Generate issues and recommendations
+        self._generate_issues(result)
+        self._generate_recommendations(result)
+
+        logger.info(f"Backlink audit complete for {domain}")
+        return result
+
+    # -- Helpers -------------------------------------------------------------
+
+    @staticmethod
+    def _is_branded_anchor(anchor: str, brand: str, domain: str) -> bool:
+        """Check if anchor text is a brand mention."""
+        brand_lower = brand.lower()
+        domain_clean = domain.replace("www.", "").split(".")[0].lower()
+        return (
+            brand_lower in anchor
+            or domain_clean in anchor
+            or domain.lower() in anchor
+        )
+
+    @staticmethod
+    def _is_naked_url(anchor: str, domain: str) -> bool:
+        """Check if anchor text is a naked URL."""
+        return (
+            anchor.startswith("http")
+            or anchor.startswith("www.")
+            or domain.lower() in anchor
+            and ("/" in anchor or "." in anchor)
+        )
+
+    @staticmethod
+    def _is_generic_anchor(anchor: str) -> bool:
+        """Check if anchor text is generic."""
+        return anchor.lower().strip() in GENERIC_ANCHORS
+
+    @staticmethod
+    def _is_exact_match(anchor: str, brand: str) -> bool:
+        """Check if anchor is exact-match keyword (not brand)."""
+        # Exact match if it looks like a keyword phrase (2+ words, no brand)
+        words = anchor.split()
+        return len(words) >= 2 and brand.lower() not in anchor.lower()
+
+    @staticmethod
+    def _extract_tld(domain: str) -> str:
+        """Extract TLD from domain."""
+        if not domain:
+            return ""
+        parts = domain.rsplit(".", 1)
+        return f".{parts[-1]}" if len(parts) > 1 else ""
+
+    @staticmethod
+    def _dr_bucket(dr: float) -> str:
+        """Map DR value to bucket range."""
+        if dr <= 10:
+            return "0-10"
+        elif dr <= 20:
+            return "11-20"
+        elif dr <= 30:
+            return "21-30"
+        elif dr <= 40:
+            return "31-40"
+        elif dr <= 50:
+            return "41-50"
+        elif dr <= 60:
+            return "51-60"
+        elif dr <= 70:
+            return "61-70"
+        elif dr <= 80:
+            return "71-80"
+        elif dr <= 90:
+            return "81-90"
+        else:
+            return "91-100"
+
+    @staticmethod
+    def _find_closest_point(history: list[dict], days_ago: int) -> int:
+        """Find referring domain count closest to N days ago."""
+        if not history:
+            return 0
+        target_date = (datetime.now() - timedelta(days=days_ago)).strftime("%Y-%m-%d")
+        closest = history[0]
+        for point in history:
+            if point.get("date", "") <= target_date:
+                closest = point
+        return closest.get("referring_domains", 0)
+
+    def _generate_issues(self, result: BacklinkAuditResult) -> None:
+        """Generate audit issues based on findings."""
+        issues = []
+
+        # Toxic links
+        if result.toxic_link_count > 20:
+            issues.append({
+                "type": "error",
+                "category": "toxic_links",
+                "message": f"High toxic link count: {result.toxic_link_count} toxic backlinks detected",
+            })
+        elif result.toxic_link_count > 5:
+            issues.append({
+                "type": "warning",
+                "category": "toxic_links",
+                "message": f"Moderate toxic links: {result.toxic_link_count} potentially harmful backlinks",
+            })
+
+        # Low DR
+        if result.domain_rating < 20:
+            issues.append({
+                "type": "warning",
+                "category": "domain_authority",
+                "message": f"Low Domain Rating ({result.domain_rating}) - weak backlink profile",
+            })
+
+        # Dofollow ratio
+        if result.dofollow_ratio < 0.5:
+            issues.append({
+                "type": "warning",
+                "category": "link_quality",
+                "message": f"Low dofollow ratio ({result.dofollow_ratio:.0%}) - majority are nofollow",
+            })
+
+        # Anchor distribution issues
+        if result.anchor_distribution:
+            if result.anchor_distribution.exact_match_pct > 30:
+                issues.append({
+                    "type": "error",
+                    "category": "anchor_text",
+                    "message": f"Over-optimized anchors: {result.anchor_distribution.exact_match_pct}% exact match (risk of penalty)",
+                })
+            if result.anchor_distribution.branded_pct < 10:
+                issues.append({
+                    "type": "warning",
+                    "category": "anchor_text",
+                    "message": f"Low branded anchors ({result.anchor_distribution.branded_pct}%) - unnatural profile",
+                })
+
+        # Broken backlinks
+        if len(result.broken_backlinks) > 10:
+            issues.append({
+                "type": "warning",
+                "category": "broken_links",
+                "message": f"{len(result.broken_backlinks)} broken backlinks found - recovery opportunities",
+            })
+
+        # Velocity issues
+        if result.link_velocity and result.link_velocity.velocity_trend == "declining":
+            issues.append({
+                "type": "warning",
+                "category": "velocity",
+                "message": "Declining link velocity - losing more links than gaining",
+            })
+
+        result.issues = issues
+
+    def _generate_recommendations(self, result: BacklinkAuditResult) -> None:
+        """Generate actionable recommendations."""
+        recs = []
+
+        if result.toxic_risk_level in ("critical", "high"):
+            recs.append(
+                "Disavow toxic backlinks immediately using Google Disavow Tool. "
+                f"Priority: {result.toxic_link_count} toxic links detected."
+            )
+
+        if result.domain_rating < 30:
+            recs.append(
+                "Focus on acquiring high-DR backlinks (DR 40+) from authoritative "
+                "domains in your niche to improve Domain Rating."
+            )
+
+        if result.anchor_distribution and result.anchor_distribution.exact_match_pct > 25:
+            recs.append(
+                "Diversify anchor text profile. Reduce exact-match anchors and "
+                "increase branded, generic, and naked URL anchors."
+            )
+
+        if len(result.broken_backlinks) > 5:
+            high_dr_broken = [b for b in result.broken_backlinks if b.get("dr", 0) > 30]
+            recs.append(
+                f"Reclaim {len(result.broken_backlinks)} broken backlinks "
+                f"({len(high_dr_broken)} from DR 30+ domains). Set up 301 redirects "
+                "or recreate content at original URLs."
+            )
+
+        if result.korean_platforms and result.korean_platforms.total_korean < 10:
+            recs.append(
+                "Increase presence on Korean platforms (Naver Blog, Tistory, Brunch). "
+                "Korean platform links signal local relevance for Korean SERP rankings."
+            )
+
+        if result.link_velocity and result.link_velocity.velocity_trend == "declining":
+            recs.append(
+                "Reverse declining link velocity with active outreach: guest posting, "
+                "digital PR, and content promotion campaigns."
+            )
+
+        if result.edu_gov_backlinks == 0:
+            recs.append(
+                "Target .edu and .gov backlinks through scholarship programs, "
+                "research collaborations, or government resource pages."
+            )
+
+        result.recommendations = recs
+
+
+# ---------------------------------------------------------------------------
+# Output Formatting
+# ---------------------------------------------------------------------------
+
+
+def format_rich_output(result: BacklinkAuditResult) -> None:
+    """Display audit results using Rich tables."""
+    console.print(f"\n[bold cyan]Backlink Audit: {result.domain}[/bold cyan]")
+    console.print(f"[dim]Timestamp: {result.timestamp}[/dim]\n")
+
+    # Overview table
+    overview = Table(title="Overview", show_header=True, header_style="bold magenta")
+    overview.add_column("Metric", style="cyan")
+    overview.add_column("Value", style="green")
+    overview.add_row("Domain Rating", str(result.domain_rating))
+    overview.add_row("Total Backlinks", f"{result.total_backlinks:,}")
+    overview.add_row("Referring Domains", f"{result.referring_domains:,}")
+    overview.add_row("Dofollow Ratio", f"{result.dofollow_ratio:.0%}")
+    overview.add_row("Edu/Gov Backlinks", str(result.edu_gov_backlinks))
+    overview.add_row("Toxic Links", f"{result.toxic_link_count} ({result.toxic_risk_level})")
+    console.print(overview)
+
+    # Anchor distribution
+    if result.anchor_distribution:
+        anchor_table = Table(title="\nAnchor Distribution", show_header=True, header_style="bold magenta")
+        anchor_table.add_column("Type", style="cyan")
+        anchor_table.add_column("Count", style="green")
+        anchor_table.add_column("Percentage", style="yellow")
+        ad = result.anchor_distribution
+        anchor_table.add_row("Branded", str(ad.branded), f"{ad.branded_pct}%")
+        anchor_table.add_row("Exact Match", str(ad.exact_match), f"{ad.exact_match_pct}%")
+        anchor_table.add_row("Partial Match", str(ad.partial_match), f"{(ad.partial_match / max(ad.total, 1) * 100):.1f}%")
+        anchor_table.add_row("Generic", str(ad.generic), f"{ad.generic_pct}%")
+        anchor_table.add_row("Naked URL", str(ad.naked_url), f"{ad.naked_url_pct}%")
+        console.print(anchor_table)
+
+    # DR distribution
+    if result.dr_distribution:
+        dr_table = Table(title="\nDR Distribution (Referring Domains)", show_header=True, header_style="bold magenta")
+        dr_table.add_column("DR Range", style="cyan")
+        dr_table.add_column("Count", style="green")
+        for bucket, count in result.dr_distribution.items():
+            dr_table.add_row(bucket, str(count))
+        console.print(dr_table)
+
+    # Toxic links (top 10)
+    if result.toxic_links:
+        toxic_table = Table(title=f"\nToxic Links (Top 10 of {len(result.toxic_links)})", show_header=True, header_style="bold red")
+        toxic_table.add_column("Domain", style="red")
+        toxic_table.add_column("Risk", style="yellow")
+        toxic_table.add_column("Reason", style="dim")
+        for tl in result.toxic_links[:10]:
+            toxic_table.add_row(tl.domain, f"{tl.risk_score:.0f}", tl.reason[:60])
+        console.print(toxic_table)
+
+    # Korean platforms
+    if result.korean_platforms and result.korean_platforms.total_korean > 0:
+        kr_table = Table(title="\nKorean Platform Links", show_header=True, header_style="bold magenta")
+        kr_table.add_column("Platform", style="cyan")
+        kr_table.add_column("Count", style="green")
+        kp = result.korean_platforms
+        for plat, val in [
+            ("Naver Blog", kp.naver_blog), ("Naver Cafe", kp.naver_cafe),
+            ("Naver Post", kp.naver_post), ("Tistory", kp.tistory),
+            ("Brunch", kp.brunch), ("Daum Blog", kp.daum_blog),
+            ("Korean News", kp.korean_news), ("Korean Community", kp.korean_community),
+        ]:
+            if val > 0:
+                kr_table.add_row(plat, str(val))
+        kr_table.add_row("[bold]Total Korean[/bold]", f"[bold]{kp.total_korean}[/bold]")
+        console.print(kr_table)
+
+    # Link velocity
+    if result.link_velocity:
+        vel_table = Table(title="\nLink Velocity", show_header=True, header_style="bold magenta")
+        vel_table.add_column("Period", style="cyan")
+        vel_table.add_column("New", style="green")
+        vel_table.add_column("Lost", style="red")
+        lv = result.link_velocity
+        vel_table.add_row("Last 7 days", str(lv.new_last_7d), str(lv.lost_last_7d))
+        vel_table.add_row("Last 30 days", str(lv.new_last_30d), str(lv.lost_last_30d))
+        vel_table.add_row("Last 90 days", str(lv.new_last_90d), str(lv.lost_last_90d))
+        vel_table.add_row("Trend", lv.velocity_trend, "")
+        console.print(vel_table)
+
+    # Broken backlinks
+    if result.broken_backlinks:
+        broken_table = Table(title=f"\nBroken Backlinks (Top 10 of {len(result.broken_backlinks)})", show_header=True, header_style="bold magenta")
+        broken_table.add_column("Source Domain", style="cyan")
+        broken_table.add_column("DR", style="green")
+        broken_table.add_column("Target URL", style="dim")
+        for bl in result.broken_backlinks[:10]:
+            broken_table.add_row(
+                bl.get("source_domain", ""),
+                str(bl.get("dr", 0)),
+                bl.get("target_url", "")[:50],
+            )
+        console.print(broken_table)
+
+    # Issues
+    if result.issues:
+        console.print("\n[bold red]Issues Found:[/bold red]")
+        for issue in result.issues:
+            icon = "[red]ERROR[/red]" if issue["type"] == "error" else "[yellow]WARN[/yellow]"
+            console.print(f"  {icon} [{issue['category']}] {issue['message']}")
+
+    # Recommendations
+    if result.recommendations:
+        console.print("\n[bold green]Recommendations:[/bold green]")
+        for i, rec in enumerate(result.recommendations, 1):
+            console.print(f"  {i}. {rec}")
+
+    console.print()
+
+
+def result_to_dict(result: BacklinkAuditResult) -> dict[str, Any]:
+    """Convert audit result to JSON-serializable dict."""
+    data = {
+        "url": result.url,
+        "domain": result.domain,
+        "domain_rating": result.domain_rating,
+        "backlink_stats": {
+            "total_backlinks": result.total_backlinks,
+            "referring_domains": result.referring_domains,
+            "dofollow_backlinks": result.dofollow_backlinks,
+            "nofollow_backlinks": result.nofollow_backlinks,
+            "dofollow_ratio": result.dofollow_ratio,
+            "edu_gov_backlinks": result.edu_gov_backlinks,
+        },
+        "anchor_distribution": asdict(result.anchor_distribution) if result.anchor_distribution else None,
+        "dr_distribution": result.dr_distribution,
+        "country_distribution": result.country_distribution,
+        "toxic_links": [asdict(t) for t in result.toxic_links],
+        "toxic_link_count": result.toxic_link_count,
+        "toxic_risk_level": result.toxic_risk_level,
+        "broken_backlinks": result.broken_backlinks,
+        "korean_platforms": asdict(result.korean_platforms) if result.korean_platforms else None,
+        "link_velocity": asdict(result.link_velocity) if result.link_velocity else None,
+        "issues": result.issues,
+        "recommendations": result.recommendations,
+        "timestamp": result.timestamp,
+    }
+    return data
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+
+def parse_args() -> argparse.Namespace:
+    """Parse command-line arguments."""
+    parser = argparse.ArgumentParser(
+        description="Backlink Auditor - Analyze backlink profiles and detect toxic links",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python backlink_auditor.py --url https://example.com --json
+  python backlink_auditor.py --url https://example.com --velocity --broken --json
+  python backlink_auditor.py --url https://example.com --korean-platforms --json
+  python backlink_auditor.py --url https://example.com --velocity --broken --korean-platforms --output report.json
+        """,
+    )
+    parser.add_argument("--url", required=True, help="Target URL or domain to audit")
+    parser.add_argument("--brand", default="", help="Brand name for anchor classification")
+    parser.add_argument("--velocity", action="store_true", help="Include link velocity tracking")
+    parser.add_argument("--broken", action="store_true", help="Include broken backlink analysis")
+    parser.add_argument("--korean-platforms", action="store_true", help="Include Korean platform link mapping")
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", "-o", help="Save output to file")
+    return parser.parse_args()
+
+
+async def main() -> None:
+    """Main entry point."""
+    args = parse_args()
+
+    auditor = BacklinkAuditor()
+
+    try:
+        result = await auditor.audit(
+            url=args.url,
+            include_velocity=args.velocity,
+            include_broken=args.broken,
+            include_korean=args.korean_platforms,
+            brand_name=args.brand,
+        )
+
+        if args.json or args.output:
+            output_data = result_to_dict(result)
+            json_str = json.dumps(output_data, indent=2, ensure_ascii=False)
+
+            if args.output:
+                with open(args.output, "w", encoding="utf-8") as f:
+                    f.write(json_str)
+                logger.info(f"Report saved to {args.output}")
+
+            if args.json:
+                print(json_str)
+        else:
+            format_rich_output(result)
+
+        auditor.print_stats()
+
+    except KeyboardInterrupt:
+        logger.warning("Audit interrupted by user")
+        sys.exit(1)
+    except Exception as e:
+        logger.error(f"Audit failed: {e}")
+        if args.json:
+            print(json.dumps({"error": str(e)}, indent=2))
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/22-seo-link-building/code/scripts/base_client.py b/custom-skills/22-seo-link-building/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/22-seo-link-building/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/22-seo-link-building/code/scripts/link_gap_finder.py b/custom-skills/22-seo-link-building/code/scripts/link_gap_finder.py
new file mode 100644
index 0000000..9264d0f
--- /dev/null
+++ b/custom-skills/22-seo-link-building/code/scripts/link_gap_finder.py
@@ -0,0 +1,802 @@
+"""
+Link Gap Finder - Competitor link gap analysis
+===============================================
+Purpose: Identify link building opportunities by finding domains that link
+         to competitors but not to the target site via Ahrefs MCP.
+Python: 3.10+
+Usage:
+    python link_gap_finder.py --target https://example.com --competitor https://comp1.com --json
+    python link_gap_finder.py --target https://example.com --competitor https://comp1.com --competitor https://comp2.com --min-dr 30 --json
+"""
+
+from __future__ import annotations
+
+import argparse
+import asyncio
+import json
+import logging
+import re
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from typing import Any
+from urllib.parse import urlparse
+
+import aiohttp
+import pandas as pd
+from rich.console import Console
+from rich.table import Table
+
+from base_client import BaseAsyncClient, config
+
+# ---------------------------------------------------------------------------
+# Logging
+# ---------------------------------------------------------------------------
+logger = logging.getLogger("link_gap_finder")
+console = Console()
+
+# ---------------------------------------------------------------------------
+# Constants
+# ---------------------------------------------------------------------------
+AHREFS_BASE = "https://api.ahrefs.com/v3"
+
+# Source category detection patterns
+SOURCE_CATEGORY_PATTERNS: dict[str, list[str]] = {
+    "news": [
+        "news", "press", "media", "journal", "herald", "times", "post",
+        "gazette", "tribune", "daily", "chosun", "donga", "joongang",
+        "hani", "khan", "yna", "yonhap", "reuters", "bloomberg",
+        "techcrunch", "verge", "wired", "arstechnica", "bbc", "cnn",
+    ],
+    "blog": [
+        "blog", "wordpress", "medium.com", "tistory.com", "brunch.co.kr",
+        "blog.naver.com", "tumblr", "blogger", "substack", "ghost.io",
+        "velog.io", "dev.to",
+    ],
+    "forum": [
+        "forum", "community", "discuss", "reddit.com", "quora.com",
+        "stackexchange", "stackoverflow", "cafe.naver.com", "dcinside",
+        "fmkorea", "clien", "ppomppu", "theqoo", "ruliweb",
+    ],
+    "directory": [
+        "directory", "listing", "yellowpages", "yelp", "bbb.org",
+        "clutch.co", "g2.com", "capterra", "trustpilot", "glassdoor",
+        "dmoz", "aboutus", "hotfrog", "manta", "superpages",
+    ],
+    "edu_gov": [
+        ".edu", ".gov", ".ac.kr", ".go.kr", ".or.kr",
+    ],
+    "social": [
+        "facebook.com", "twitter.com", "x.com", "linkedin.com",
+        "instagram.com", "youtube.com", "pinterest.com", "tiktok.com",
+    ],
+    "korean_platform": [
+        "naver.com", "daum.net", "kakao.com", "tistory.com",
+        "brunch.co.kr", "zum.com", "nate.com",
+    ],
+}
+
+# ---------------------------------------------------------------------------
+# Dataclasses
+# ---------------------------------------------------------------------------
+
+
+@dataclass
+class LinkOpportunity:
+    """A single link building opportunity from gap analysis."""
+    domain: str
+    dr: float = 0.0
+    traffic: int = 0
+    linked_competitors: list[str] = field(default_factory=list)
+    competitor_count: int = 0
+    not_linked_target: bool = True
+    category: str = "other"
+    feasibility_score: float = 0.0
+    impact_score: float = 0.0
+    overall_score: float = 0.0
+    backlinks_to_competitors: int = 0
+    country: str = ""
+    top_anchor: str = ""
+
+
+@dataclass
+class GapSummary:
+    """Summary statistics for the gap analysis."""
+    total_opportunities: int = 0
+    avg_dr: float = 0.0
+    high_dr_count: int = 0
+    category_breakdown: dict[str, int] = field(default_factory=dict)
+    top_countries: list[dict[str, Any]] = field(default_factory=list)
+    total_competitor_refdomains: dict[str, int] = field(default_factory=dict)
+    target_refdomains_count: int = 0
+
+
+@dataclass
+class LinkGapResult:
+    """Complete link gap analysis result."""
+    target_url: str
+    target_domain: str = ""
+    competitor_urls: list[str] = field(default_factory=list)
+    competitor_domains: list[str] = field(default_factory=list)
+    target_dr: float = 0.0
+    opportunities: list[LinkOpportunity] = field(default_factory=list)
+    summary: GapSummary | None = None
+    top_opportunities: list[LinkOpportunity] = field(default_factory=list)
+    issues: list[dict[str, str]] = field(default_factory=list)
+    recommendations: list[str] = field(default_factory=list)
+    timestamp: str = ""
+
+
+# ---------------------------------------------------------------------------
+# LinkGapFinder
+# ---------------------------------------------------------------------------
+
+
+class LinkGapFinder(BaseAsyncClient):
+    """Find link building opportunities by analyzing competitor backlink gaps."""
+
+    def __init__(self, **kwargs):
+        super().__init__(max_concurrent=5, requests_per_second=2.0, **kwargs)
+        self.session: aiohttp.ClientSession | None = None
+
+    # -- Ahrefs MCP helper ---------------------------------------------------
+
+    async def _call_ahrefs(
+        self, endpoint: str, params: dict[str, Any]
+    ) -> dict[str, Any]:
+        """
+        Call Ahrefs API endpoint.
+
+        In MCP context this calls mcp__ahrefs__<endpoint>.
+        For standalone use, falls back to REST API with token.
+        """
+        api_token = config.get_required("AHREFS_API_TOKEN") if not self.session else None
+
+        if self.session and api_token:
+            url = f"{AHREFS_BASE}/{endpoint}"
+            headers = {"Authorization": f"Bearer {api_token}"}
+            async with self.session.get(url, headers=headers, params=params) as resp:
+                resp.raise_for_status()
+                return await resp.json()
+
+        logger.warning(
+            f"Ahrefs call to '{endpoint}' - use MCP tool "
+            f"mcp__ahrefs__{endpoint.replace('-', '_')} in Claude Desktop"
+        )
+        return {"endpoint": endpoint, "params": params, "data": [], "note": "mcp_stub"}
+
+    # -- Core methods --------------------------------------------------------
+
+    async def get_referring_domains(
+        self, url: str, limit: int = 1000
+    ) -> list[dict[str, Any]]:
+        """Fetch referring domains for a given URL/domain."""
+        target = urlparse(url).netloc or url
+        result = await self._call_ahrefs(
+            "site-explorer-referring-domains",
+            {"target": target, "mode": "domain", "limit": limit, "order_by": "domain_rating:desc"},
+        )
+        domains = result.get("data", result.get("refdomains", []))
+        if isinstance(domains, dict):
+            domains = domains.get("refdomains", [])
+        return domains if isinstance(domains, list) else []
+
+    async def get_domain_rating(self, url: str) -> float:
+        """Fetch Domain Rating for a URL."""
+        target = urlparse(url).netloc or url
+        result = await self._call_ahrefs(
+            "site-explorer-domain-rating",
+            {"target": target},
+        )
+        data = result.get("data", result) if isinstance(result, dict) else {}
+        return data.get("domain_rating", 0.0)
+
+    async def get_domain_metrics(self, url: str) -> dict[str, Any]:
+        """Fetch comprehensive domain metrics."""
+        target = urlparse(url).netloc or url
+        result = await self._call_ahrefs(
+            "site-explorer-backlinks-stats",
+            {"target": target, "mode": "domain"},
+        )
+        data = result.get("data", result) if isinstance(result, dict) else {}
+        return {
+            "total_backlinks": data.get("live", 0),
+            "referring_domains": data.get("live_refdomains", 0),
+            "dofollow": data.get("live_dofollow", 0),
+        }
+
+    def find_gaps(
+        self,
+        target_domains: set[str],
+        competitor_domain_maps: dict[str, set[str]],
+    ) -> list[dict[str, Any]]:
+        """
+        Find domains linking to competitors but not to the target.
+
+        Returns a list of gap domains with metadata about which
+        competitors they link to.
+        """
+        # Collect all competitor referring domains
+        all_competitor_domains: dict[str, list[str]] = {}
+
+        for comp_name, comp_domains in competitor_domain_maps.items():
+            for domain in comp_domains:
+                domain_lower = domain.lower()
+                if domain_lower not in all_competitor_domains:
+                    all_competitor_domains[domain_lower] = []
+                all_competitor_domains[domain_lower].append(comp_name)
+
+        # Find gaps: in competitor set but not in target set
+        target_set_lower = {d.lower() for d in target_domains}
+        gaps = []
+
+        for domain, linked_comps in all_competitor_domains.items():
+            if domain not in target_set_lower:
+                gaps.append({
+                    "domain": domain,
+                    "linked_competitors": linked_comps,
+                    "competitor_count": len(set(linked_comps)),
+                })
+
+        # Sort by number of competitors linking (more = higher priority)
+        gaps.sort(key=lambda g: g["competitor_count"], reverse=True)
+        return gaps
+
+    def score_opportunities(
+        self,
+        gaps: list[dict[str, Any]],
+        refdomains_data: dict[str, list[dict[str, Any]]],
+        total_competitors: int,
+    ) -> list[LinkOpportunity]:
+        """
+        Score gap opportunities by DR, traffic, relevance, and feasibility.
+
+        Scoring factors:
+        - DR weight: Higher DR = more impactful link
+        - Competitor overlap: More competitors linking = easier to acquire
+        - Category bonus: Editorial/news links valued higher
+        - Traffic bonus: Higher traffic domains valued more
+        """
+        # Build a lookup of domain metadata from competitor refdomains
+        domain_metadata: dict[str, dict[str, Any]] = {}
+        for comp_url, domains in refdomains_data.items():
+            for rd in domains:
+                d = rd.get("domain", rd.get("domain_from", "")).lower()
+                if d and d not in domain_metadata:
+                    domain_metadata[d] = {
+                        "dr": rd.get("domain_rating", rd.get("dr", 0)),
+                        "traffic": rd.get("organic_traffic", rd.get("traffic", 0)),
+                        "backlinks": rd.get("backlinks", 0),
+                        "country": rd.get("country", ""),
+                    }
+
+        opportunities = []
+
+        for gap in gaps:
+            domain = gap["domain"]
+            meta = domain_metadata.get(domain, {})
+
+            dr = meta.get("dr", 0)
+            traffic = meta.get("traffic", 0)
+            comp_count = gap["competitor_count"]
+
+            # Category detection
+            category = self._detect_category(domain)
+
+            # Feasibility score (0-100)
+            # Higher if: more competitors link (social proof), blog/forum (easier outreach)
+            feasibility = min(100, (
+                (comp_count / max(total_competitors, 1)) * 40  # Competitor overlap
+                + (30 if category in ("blog", "forum", "directory") else 10)  # Category ease
+                + (20 if dr < 60 else 5)  # Lower DR = easier to get link from
+                + (10 if traffic > 0 else 0)  # Active site bonus
+            ))
+
+            # Impact score (0-100)
+            # Higher if: high DR, high traffic, editorial/news
+            impact = min(100, (
+                min(dr, 100) * 0.4  # DR weight (40%)
+                + min(traffic / 1000, 30)  # Traffic weight (up to 30)
+                + (20 if category in ("news", "edu_gov") else 5)  # Authority bonus
+                + (comp_count / max(total_competitors, 1)) * 10  # Validation
+            ))
+
+            # Overall score = weighted average
+            overall = round(feasibility * 0.4 + impact * 0.6, 1)
+
+            opp = LinkOpportunity(
+                domain=domain,
+                dr=dr,
+                traffic=traffic,
+                linked_competitors=gap["linked_competitors"],
+                competitor_count=comp_count,
+                not_linked_target=True,
+                category=category,
+                feasibility_score=round(feasibility, 1),
+                impact_score=round(impact, 1),
+                overall_score=overall,
+                backlinks_to_competitors=meta.get("backlinks", 0),
+                country=meta.get("country", ""),
+            )
+            opportunities.append(opp)
+
+        # Sort by overall score descending
+        opportunities.sort(key=lambda o: o.overall_score, reverse=True)
+        return opportunities
+
+    def categorize_sources(
+        self, opportunities: list[LinkOpportunity]
+    ) -> dict[str, list[LinkOpportunity]]:
+        """Group opportunities by source category."""
+        categorized: dict[str, list[LinkOpportunity]] = {}
+        for opp in opportunities:
+            cat = opp.category
+            if cat not in categorized:
+                categorized[cat] = []
+            categorized[cat].append(opp)
+        return categorized
+
+    # -- Orchestration -------------------------------------------------------
+
+    async def analyze(
+        self,
+        target_url: str,
+        competitor_urls: list[str],
+        min_dr: float = 0,
+        country_filter: str = "",
+        limit: int = 1000,
+    ) -> LinkGapResult:
+        """Orchestrate full link gap analysis."""
+        target_domain = urlparse(target_url).netloc or target_url
+        comp_domains = [urlparse(c).netloc or c for c in competitor_urls]
+
+        logger.info(f"Starting link gap analysis: {target_domain} vs {comp_domains}")
+
+        result = LinkGapResult(
+            target_url=target_url,
+            target_domain=target_domain,
+            competitor_urls=competitor_urls,
+            competitor_domains=comp_domains,
+            timestamp=datetime.now().isoformat(),
+        )
+
+        # Phase 1: Fetch target DR and referring domains
+        logger.info("Phase 1: Fetching target data...")
+        target_dr_task = self.get_domain_rating(target_url)
+        target_rd_task = self.get_referring_domains(target_url, limit=limit)
+
+        target_dr, target_refdomains = await asyncio.gather(
+            target_dr_task, target_rd_task, return_exceptions=True,
+        )
+
+        result.target_dr = target_dr if isinstance(target_dr, (int, float)) else 0
+        target_rd_list = target_refdomains if isinstance(target_refdomains, list) else []
+        target_domain_set = {
+            rd.get("domain", rd.get("domain_from", "")).lower()
+            for rd in target_rd_list
+            if rd.get("domain", rd.get("domain_from", ""))
+        }
+
+        # Phase 2: Fetch competitor referring domains (parallel)
+        logger.info("Phase 2: Fetching competitor data...")
+        comp_rd_tasks = {
+            comp_url: self.get_referring_domains(comp_url, limit=limit)
+            for comp_url in competitor_urls
+        }
+        comp_results = {}
+        for comp_url, task in comp_rd_tasks.items():
+            try:
+                comp_rd = await task
+                comp_results[comp_url] = comp_rd if isinstance(comp_rd, list) else []
+            except Exception as e:
+                logger.error(f"Failed to fetch refdomains for {comp_url}: {e}")
+                comp_results[comp_url] = []
+
+        # Build competitor domain maps
+        competitor_domain_maps: dict[str, set[str]] = {}
+        for comp_url, rd_list in comp_results.items():
+            comp_domain = urlparse(comp_url).netloc or comp_url
+            competitor_domain_maps[comp_domain] = {
+                rd.get("domain", rd.get("domain_from", "")).lower()
+                for rd in rd_list
+                if rd.get("domain", rd.get("domain_from", ""))
+            }
+
+        # Phase 3: Find gaps
+        logger.info("Phase 3: Finding link gaps...")
+        raw_gaps = self.find_gaps(target_domain_set, competitor_domain_maps)
+        logger.info(f"Found {len(raw_gaps)} gap domains")
+
+        # Phase 4: Score opportunities
+        logger.info("Phase 4: Scoring opportunities...")
+        opportunities = self.score_opportunities(
+            raw_gaps, comp_results, len(competitor_urls)
+        )
+
+        # Apply filters
+        if min_dr > 0:
+            opportunities = [o for o in opportunities if o.dr >= min_dr]
+
+        if country_filter:
+            country_lower = country_filter.lower()
+            opportunities = [
+                o for o in opportunities
+                if o.country.lower() == country_lower or not o.country
+            ]
+
+        result.opportunities = opportunities
+        result.top_opportunities = opportunities[:50]
+
+        # Phase 5: Build summary
+        logger.info("Phase 5: Building summary...")
+        result.summary = self._build_summary(
+            opportunities, comp_results, len(target_rd_list)
+        )
+
+        # Phase 6: Generate recommendations
+        self._generate_issues(result)
+        self._generate_recommendations(result)
+
+        logger.info(f"Link gap analysis complete: {len(opportunities)} opportunities found")
+        return result
+
+    # -- Helpers -------------------------------------------------------------
+
+    @staticmethod
+    def _detect_category(domain: str) -> str:
+        """Detect the category of a domain based on patterns."""
+        domain_lower = domain.lower()
+
+        for category, patterns in SOURCE_CATEGORY_PATTERNS.items():
+            for pattern in patterns:
+                if pattern in domain_lower:
+                    return category
+
+        # Fallback heuristics
+        if domain_lower.endswith((".edu", ".ac.kr", ".gov", ".go.kr")):
+            return "edu_gov"
+
+        return "other"
+
+    def _build_summary(
+        self,
+        opportunities: list[LinkOpportunity],
+        comp_results: dict[str, list],
+        target_rd_count: int,
+    ) -> GapSummary:
+        """Build summary statistics from opportunities."""
+        summary = GapSummary()
+        summary.total_opportunities = len(opportunities)
+        summary.target_refdomains_count = target_rd_count
+
+        if opportunities:
+            dr_values = [o.dr for o in opportunities if o.dr > 0]
+            summary.avg_dr = round(sum(dr_values) / max(len(dr_values), 1), 1)
+            summary.high_dr_count = sum(1 for o in opportunities if o.dr >= 50)
+
+            # Category breakdown
+            cat_counts: dict[str, int] = {}
+            country_counts: dict[str, int] = {}
+            for opp in opportunities:
+                cat_counts[opp.category] = cat_counts.get(opp.category, 0) + 1
+                if opp.country:
+                    country_counts[opp.country] = country_counts.get(opp.country, 0) + 1
+
+            summary.category_breakdown = dict(
+                sorted(cat_counts.items(), key=lambda x: x[1], reverse=True)
+            )
+            summary.top_countries = sorted(
+                [{"country": k, "count": v} for k, v in country_counts.items()],
+                key=lambda x: x["count"], reverse=True,
+            )[:10]
+
+        # Competitor refdomains counts
+        for comp_url, rd_list in comp_results.items():
+            comp_domain = urlparse(comp_url).netloc or comp_url
+            summary.total_competitor_refdomains[comp_domain] = len(rd_list)
+
+        return summary
+
+    def _generate_issues(self, result: LinkGapResult) -> None:
+        """Generate issues based on gap analysis."""
+        issues = []
+
+        if result.summary:
+            # Large gap warning
+            if result.summary.total_opportunities > 500:
+                issues.append({
+                    "type": "warning",
+                    "category": "link_gap",
+                    "message": (
+                        f"Large link gap: {result.summary.total_opportunities} domains "
+                        "link to competitors but not to you"
+                    ),
+                })
+
+            # High-DR gap
+            if result.summary.high_dr_count > 50:
+                issues.append({
+                    "type": "error",
+                    "category": "authority_gap",
+                    "message": (
+                        f"{result.summary.high_dr_count} high-authority domains (DR 50+) "
+                        "link to competitors but not to you"
+                    ),
+                })
+
+            # Category-specific gaps
+            news_gap = result.summary.category_breakdown.get("news", 0)
+            if news_gap > 20:
+                issues.append({
+                    "type": "warning",
+                    "category": "pr_gap",
+                    "message": f"{news_gap} news/media domains link to competitors - consider digital PR",
+                })
+
+            edu_gap = result.summary.category_breakdown.get("edu_gov", 0)
+            if edu_gap > 5:
+                issues.append({
+                    "type": "info",
+                    "category": "edu_gov_gap",
+                    "message": f"{edu_gap} .edu/.gov domains link to competitors - high-authority opportunity",
+                })
+
+        result.issues = issues
+
+    def _generate_recommendations(self, result: LinkGapResult) -> None:
+        """Generate actionable recommendations."""
+        recs = []
+
+        if not result.opportunities:
+            recs.append("No significant link gaps found. Consider expanding competitor list.")
+            result.recommendations = recs
+            return
+
+        # Top opportunities by category
+        categorized = self.categorize_sources(result.top_opportunities[:100])
+
+        if "news" in categorized:
+            news_count = len(categorized["news"])
+            top_news = [o.domain for o in categorized["news"][:3]]
+            recs.append(
+                f"Pursue {news_count} news/media link opportunities. "
+                f"Top targets: {', '.join(top_news)}. "
+                "Strategy: create newsworthy content, press releases, expert commentary."
+            )
+
+        if "blog" in categorized:
+            blog_count = len(categorized["blog"])
+            recs.append(
+                f"Target {blog_count} blog/content site opportunities via guest posting, "
+                "collaborative content, and expert interviews."
+            )
+
+        if "directory" in categorized:
+            dir_count = len(categorized["directory"])
+            recs.append(
+                f"Submit to {dir_count} relevant directories and listing sites. "
+                "Low effort, moderate impact for local SEO signals."
+            )
+
+        if "forum" in categorized:
+            forum_count = len(categorized["forum"])
+            recs.append(
+                f"Engage in {forum_count} forum/community sites with helpful answers "
+                "and resource sharing. Build presence before linking."
+            )
+
+        if "korean_platform" in categorized:
+            kr_count = len(categorized["korean_platform"])
+            recs.append(
+                f"Build presence on {kr_count} Korean platforms (Naver, Tistory, Brunch). "
+                "Critical for Korean SERP visibility."
+            )
+
+        if "edu_gov" in categorized:
+            eg_count = len(categorized["edu_gov"])
+            recs.append(
+                f"Target {eg_count} .edu/.gov link opportunities through scholarship "
+                "programs, research partnerships, or government resource contributions."
+            )
+
+        # Multi-competitor overlap
+        multi_comp = [o for o in result.top_opportunities if o.competitor_count >= 2]
+        if multi_comp:
+            recs.append(
+                f"{len(multi_comp)} domains link to multiple competitors but not to you. "
+                "These are high-priority targets as they validate industry relevance."
+            )
+
+        # Quick wins: high feasibility, moderate impact
+        quick_wins = [
+            o for o in result.opportunities[:100]
+            if o.feasibility_score >= 60 and o.impact_score >= 30
+        ]
+        if quick_wins:
+            recs.append(
+                f"Prioritize {len(quick_wins)} quick-win opportunities with high "
+                "feasibility and moderate impact for fastest link acquisition."
+            )
+
+        result.recommendations = recs
+
+
+# ---------------------------------------------------------------------------
+# Output Formatting
+# ---------------------------------------------------------------------------
+
+
+def format_rich_output(result: LinkGapResult) -> None:
+    """Display gap analysis results using Rich tables."""
+    console.print(f"\n[bold cyan]Link Gap Analysis: {result.target_domain}[/bold cyan]")
+    console.print(f"[dim]vs {', '.join(result.competitor_domains)}[/dim]")
+    console.print(f"[dim]Timestamp: {result.timestamp}[/dim]\n")
+
+    # Summary
+    if result.summary:
+        summary_table = Table(title="Summary", show_header=True, header_style="bold magenta")
+        summary_table.add_column("Metric", style="cyan")
+        summary_table.add_column("Value", style="green")
+        summary_table.add_row("Target DR", str(result.target_dr))
+        summary_table.add_row("Target Referring Domains", str(result.summary.target_refdomains_count))
+        summary_table.add_row("Total Gap Opportunities", str(result.summary.total_opportunities))
+        summary_table.add_row("Avg Opportunity DR", str(result.summary.avg_dr))
+        summary_table.add_row("High-DR Opportunities (50+)", str(result.summary.high_dr_count))
+
+        for comp, count in result.summary.total_competitor_refdomains.items():
+            summary_table.add_row(f"  {comp} Refdomains", str(count))
+
+        console.print(summary_table)
+
+    # Category breakdown
+    if result.summary and result.summary.category_breakdown:
+        cat_table = Table(title="\nCategory Breakdown", show_header=True, header_style="bold magenta")
+        cat_table.add_column("Category", style="cyan")
+        cat_table.add_column("Count", style="green")
+        for cat, count in result.summary.category_breakdown.items():
+            cat_table.add_row(cat, str(count))
+        console.print(cat_table)
+
+    # Top opportunities
+    if result.top_opportunities:
+        opp_table = Table(
+            title=f"\nTop Opportunities (showing {min(25, len(result.top_opportunities))})",
+            show_header=True,
+            header_style="bold magenta",
+        )
+        opp_table.add_column("Domain", style="cyan", max_width=35)
+        opp_table.add_column("DR", style="green", justify="right")
+        opp_table.add_column("Category", style="yellow")
+        opp_table.add_column("Comps", justify="right")
+        opp_table.add_column("Score", style="bold green", justify="right")
+        opp_table.add_column("Feasibility", justify="right")
+        opp_table.add_column("Impact", justify="right")
+
+        for opp in result.top_opportunities[:25]:
+            opp_table.add_row(
+                opp.domain[:35],
+                str(int(opp.dr)),
+                opp.category,
+                str(opp.competitor_count),
+                f"{opp.overall_score:.1f}",
+                f"{opp.feasibility_score:.0f}",
+                f"{opp.impact_score:.0f}",
+            )
+        console.print(opp_table)
+
+    # Issues
+    if result.issues:
+        console.print("\n[bold red]Issues:[/bold red]")
+        for issue in result.issues:
+            icon_map = {"error": "[red]ERROR[/red]", "warning": "[yellow]WARN[/yellow]", "info": "[blue]INFO[/blue]"}
+            icon = icon_map.get(issue["type"], "[dim]INFO[/dim]")
+            console.print(f"  {icon} [{issue['category']}] {issue['message']}")
+
+    # Recommendations
+    if result.recommendations:
+        console.print("\n[bold green]Recommendations:[/bold green]")
+        for i, rec in enumerate(result.recommendations, 1):
+            console.print(f"  {i}. {rec}")
+
+    console.print()
+
+
+def result_to_dict(result: LinkGapResult) -> dict[str, Any]:
+    """Convert gap result to JSON-serializable dict."""
+    return {
+        "target_url": result.target_url,
+        "target_domain": result.target_domain,
+        "target_dr": result.target_dr,
+        "competitor_urls": result.competitor_urls,
+        "competitor_domains": result.competitor_domains,
+        "summary": asdict(result.summary) if result.summary else None,
+        "opportunities": [asdict(o) for o in result.opportunities],
+        "top_opportunities": [asdict(o) for o in result.top_opportunities],
+        "issues": result.issues,
+        "recommendations": result.recommendations,
+        "timestamp": result.timestamp,
+    }
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+
+def parse_args() -> argparse.Namespace:
+    """Parse command-line arguments."""
+    parser = argparse.ArgumentParser(
+        description="Link Gap Finder - Identify link building opportunities vs competitors",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python link_gap_finder.py --target https://example.com --competitor https://comp1.com --json
+  python link_gap_finder.py --target https://example.com --competitor https://comp1.com --competitor https://comp2.com --min-dr 30 --json
+  python link_gap_finder.py --target https://example.com --competitor https://comp1.com --country kr --output gap_report.json
+        """,
+    )
+    parser.add_argument("--target", required=True, help="Target URL or domain")
+    parser.add_argument(
+        "--competitor", action="append", required=True,
+        help="Competitor URL or domain (can be repeated)",
+    )
+    parser.add_argument(
+        "--min-dr", type=float, default=0,
+        help="Minimum DR filter for opportunities (default: 0)",
+    )
+    parser.add_argument(
+        "--country", default="",
+        help="Filter by country code (e.g., kr, us, jp)",
+    )
+    parser.add_argument(
+        "--limit", type=int, default=1000,
+        help="Max referring domains to fetch per site (default: 1000)",
+    )
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", "-o", help="Save output to file")
+    return parser.parse_args()
+
+
+async def main() -> None:
+    """Main entry point."""
+    args = parse_args()
+
+    finder = LinkGapFinder()
+
+    try:
+        result = await finder.analyze(
+            target_url=args.target,
+            competitor_urls=args.competitor,
+            min_dr=args.min_dr,
+            country_filter=args.country,
+            limit=args.limit,
+        )
+
+        if args.json or args.output:
+            output_data = result_to_dict(result)
+            json_str = json.dumps(output_data, indent=2, ensure_ascii=False)
+
+            if args.output:
+                with open(args.output, "w", encoding="utf-8") as f:
+                    f.write(json_str)
+                logger.info(f"Report saved to {args.output}")
+
+            if args.json:
+                print(json_str)
+        else:
+            format_rich_output(result)
+
+        finder.print_stats()
+
+    except KeyboardInterrupt:
+        logger.warning("Analysis interrupted by user")
+        sys.exit(1)
+    except Exception as e:
+        logger.error(f"Analysis failed: {e}")
+        if args.json:
+            print(json.dumps({"error": str(e)}, indent=2))
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/22-seo-link-building/code/scripts/requirements.txt b/custom-skills/22-seo-link-building/code/scripts/requirements.txt
new file mode 100644
index 0000000..dfb267a
--- /dev/null
+++ b/custom-skills/22-seo-link-building/code/scripts/requirements.txt
@@ -0,0 +1,8 @@
+# 22-seo-link-building dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+pandas>=2.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/22-seo-link-building/desktop/SKILL.md b/custom-skills/22-seo-link-building/desktop/SKILL.md
new file mode 100644
index 0000000..fe08e79
--- /dev/null
+++ b/custom-skills/22-seo-link-building/desktop/SKILL.md
@@ -0,0 +1,116 @@
+---
+name: seo-link-building
+description: |
+  Link building diagnosis and backlink analysis tool.
+  Triggers: backlink audit, link building, referring domains, toxic links, link gap, broken backlinks, 백링크 분석, 링크빌딩.
+---
+
+# SEO Link Building Diagnosis
+
+## Purpose
+
+Analyze backlink profiles, detect toxic links, find competitor link gaps, track link velocity, and map Korean platform links. Provides actionable link building recommendations.
+
+## Core Capabilities
+
+1. **Backlink Profile Audit** - DR, referring domains, dofollow ratio, anchor distribution
+2. **Toxic Link Detection** - PBN patterns, spam domains, link farm identification
+3. **Competitor Link Gap Analysis** - Domains linking to competitors but not target
+4. **Link Velocity Tracking** - New/lost referring domains over time
+5. **Broken Backlink Recovery** - Find and reclaim broken high-DR backlinks
+6. **Korean Platform Mapping** - Naver Blog, Cafe, Tistory, Brunch, Korean news
+
+## MCP Tool Usage
+
+### Ahrefs for Backlink Data
+```
+mcp__ahrefs__site-explorer-all-backlinks: Get all backlinks for a target
+mcp__ahrefs__site-explorer-backlinks-stats: Backlink statistics overview
+mcp__ahrefs__site-explorer-referring-domains: List referring domains
+mcp__ahrefs__site-explorer-anchors: Anchor text distribution
+mcp__ahrefs__site-explorer-broken-backlinks: Find broken backlinks
+mcp__ahrefs__site-explorer-domain-rating: Get Domain Rating
+mcp__ahrefs__site-explorer-domain-rating-history: DR trend over time
+mcp__ahrefs__site-explorer-refdomains-history: Referring domains trend
+mcp__ahrefs__site-explorer-linked-domains: Domains linked from target
+```
+
+### Notion for Report Storage
+```
+mcp__notion__notion-create-pages: Save audit report to SEO Audit Log
+mcp__notion__notion-update-page: Update existing audit entries
+```
+
+## Workflow
+
+### 1. Backlink Profile Audit
+1. Fetch Domain Rating via `site-explorer-domain-rating`
+2. Get backlink stats via `site-explorer-backlinks-stats`
+3. Retrieve referring domains via `site-explorer-referring-domains`
+4. Analyze anchor distribution via `site-explorer-anchors`
+5. Detect toxic links (PBN patterns, spam keywords, suspicious TLDs)
+6. Map Korean platform links from referring domains
+7. Report with issues and recommendations
+
+### 2. Link Gap Analysis
+1. Fetch target referring domains
+2. Fetch competitor referring domains (parallel)
+3. Compute set difference (competitor - target)
+4. Score opportunities by DR, traffic, category
+5. Categorize sources (news, blog, forum, directory, Korean platform)
+6. Rank by feasibility and impact
+7. Report top opportunities with recommendations
+
+### 3. Link Velocity Check
+1. Fetch refdomains-history for last 90 days
+2. Calculate new/lost referring domains per period
+3. Determine velocity trend (growing/stable/declining)
+4. Flag declining velocity as issue
+
+### 4. Broken Backlink Recovery
+1. Fetch broken backlinks via `site-explorer-broken-backlinks`
+2. Sort by DR (highest value first)
+3. Recommend 301 redirects or content recreation
+
+## Output Format
+
+```markdown
+## Link Building Audit: [domain]
+
+### Overview
+- Domain Rating: [DR]
+- Referring Domains: [count]
+- Dofollow Ratio: [ratio]
+- Toxic Links: [count] ([risk level])
+
+### Anchor Distribution
+| Type | Count | % |
+|------|-------|---|
+| Branded | [n] | [%] |
+| Exact Match | [n] | [%] |
+| Generic | [n] | [%] |
+| Naked URL | [n] | [%] |
+
+### Toxic Links (Top 10)
+| Domain | Risk Score | Reason |
+|--------|-----------|--------|
+
+### Korean Platform Links
+| Platform | Count |
+|----------|-------|
+
+### Link Velocity
+| Period | New | Lost |
+|--------|-----|------|
+
+### Recommendations
+1. [Priority actions]
+```
+
+## Notion Output (Required)
+
+All audit reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category (Link Building), Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: LINK-YYYYMMDD-NNN
diff --git a/custom-skills/22-seo-link-building/desktop/skill.yaml b/custom-skills/22-seo-link-building/desktop/skill.yaml
new file mode 100644
index 0000000..fd238a4
--- /dev/null
+++ b/custom-skills/22-seo-link-building/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-link-building
+description: |
+  Link building diagnosis and backlink analysis. Triggers: backlink audit, link building, referring domains, toxic links, link gap, broken backlinks.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/22-seo-link-building/desktop/tools/ahrefs.md b/custom-skills/22-seo-link-building/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..68f382b
--- /dev/null
+++ b/custom-skills/22-seo-link-building/desktop/tools/ahrefs.md
@@ -0,0 +1,70 @@
+# Ahrefs
+
+## Tools Used
+
+### site-explorer-all-backlinks
+- **Purpose**: Get all backlinks for a target domain
+- **Parameters**: target, mode (domain/prefix/url), limit, order_by
+- **Returns**: List of backlinks with source URL, domain, DR, anchor, dofollow status
+
+### site-explorer-backlinks-stats
+- **Purpose**: Backlink statistics overview
+- **Parameters**: target, mode
+- **Returns**: Total backlinks, referring domains, dofollow/nofollow counts
+
+### site-explorer-referring-domains
+- **Purpose**: List all referring domains
+- **Parameters**: target, mode, limit, order_by
+- **Returns**: Domains with DR, backlinks count, traffic, country
+
+### site-explorer-anchors
+- **Purpose**: Anchor text distribution
+- **Parameters**: target, mode, limit, order_by
+- **Returns**: Anchor texts with backlink and referring domain counts
+
+### site-explorer-broken-backlinks
+- **Purpose**: Find broken backlinks for recovery
+- **Parameters**: target, mode, limit, order_by
+- **Returns**: Broken links with source, target URL, HTTP code, DR
+
+### site-explorer-domain-rating
+- **Purpose**: Get Domain Rating for a target
+- **Parameters**: target
+- **Returns**: Domain Rating value and Ahrefs rank
+
+### site-explorer-domain-rating-history
+- **Purpose**: DR trend over time
+- **Parameters**: target, date_from
+- **Returns**: Historical DR data points
+
+### site-explorer-refdomains-history
+- **Purpose**: Referring domains trend over time
+- **Parameters**: target, mode, date_from
+- **Returns**: Historical referring domain counts
+
+### site-explorer-linked-domains
+- **Purpose**: Domains linked from the target
+- **Parameters**: target, mode, limit
+- **Returns**: Outgoing linked domains with counts
+
+## Configuration
+
+- Ahrefs MCP tools are available via `mcp__ahrefs__*` prefix
+- No API key needed when using MCP (handled by tool server)
+- Rate limits: Follow Ahrefs plan limits (typically 500 rows/request)
+
+## Examples
+
+```
+# Get backlink stats
+mcp__ahrefs__site-explorer-backlinks-stats(target="example.com", mode="domain")
+
+# Get referring domains sorted by DR
+mcp__ahrefs__site-explorer-referring-domains(target="example.com", mode="domain", limit=500, order_by="domain_rating:desc")
+
+# Get anchor text distribution
+mcp__ahrefs__site-explorer-anchors(target="example.com", mode="domain", limit=200)
+
+# Find broken backlinks
+mcp__ahrefs__site-explorer-broken-backlinks(target="example.com", mode="domain", limit=100)
+```
diff --git a/custom-skills/22-seo-link-building/desktop/tools/notion.md b/custom-skills/22-seo-link-building/desktop/tools/notion.md
new file mode 100644
index 0000000..fbd6d33
--- /dev/null
+++ b/custom-skills/22-seo-link-building/desktop/tools/notion.md
@@ -0,0 +1,39 @@
+# Notion
+
+## Tools Used
+
+### notion-create-pages
+- **Purpose**: Save link building audit reports to SEO Audit Log
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Required Properties**:
+  - Issue (title): Report title in Korean with date
+  - Site (url): Audited website URL
+  - Category (select): "Link Building"
+  - Priority (select): Critical / High / Medium / Low
+  - Found Date (date): YYYY-MM-DD
+  - Audit ID (rich_text): LINK-YYYYMMDD-NNN
+
+### notion-update-page
+- **Purpose**: Update existing audit entries with follow-up findings
+
+## Configuration
+
+- Notion MCP tools available via `mcp__notion__*` prefix
+- Authentication handled by MCP tool server
+
+## Examples
+
+```
+# Create a link building audit report
+mcp__notion__notion-create-pages(
+  parent={"database_id": "2c8581e5-8a1e-8035-880b-e38cefc2f3ef"},
+  properties={
+    "Issue": {"title": [{"text": {"content": "백링크 프로필 분석 - example.com (2025-01-15)"}}]},
+    "Site": {"url": "https://example.com"},
+    "Category": {"select": {"name": "Link Building"}},
+    "Priority": {"select": {"name": "High"}},
+    "Found Date": {"date": {"start": "2025-01-15"}},
+    "Audit ID": {"rich_text": [{"text": {"content": "LINK-20250115-001"}}]}
+  }
+)
+```
diff --git a/custom-skills/22-seo-link-building/desktop/tools/websearch.md b/custom-skills/22-seo-link-building/desktop/tools/websearch.md
new file mode 100644
index 0000000..d6cb004
--- /dev/null
+++ b/custom-skills/22-seo-link-building/desktop/tools/websearch.md
@@ -0,0 +1,24 @@
+# WebSearch
+
+## Tools Used
+
+### WebSearch
+- **Purpose**: Research link building strategies, competitor insights, and industry best practices
+- **Usage**: Supplement Ahrefs data with web research for context
+
+### WebFetch
+- **Purpose**: Fetch specific web pages for content analysis and link prospecting
+- **Usage**: Verify link opportunities, check page content relevance
+
+## Examples
+
+```
+# Research link building strategies for a niche
+WebSearch("link building strategies for SaaS companies 2025")
+
+# Research Korean link building opportunities
+WebSearch("네이버 블로그 백링크 전략 2025")
+
+# Check if a target page is relevant for outreach
+WebFetch("https://example.com/resources", "What topics does this page cover?")
+```
diff --git a/custom-skills/23-seo-content-strategy/code/CLAUDE.md b/custom-skills/23-seo-content-strategy/code/CLAUDE.md
new file mode 100644
index 0000000..bbb803d
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/code/CLAUDE.md
@@ -0,0 +1,142 @@
+# CLAUDE.md
+
+## Overview
+
+Content strategy tool for SEO-driven content planning. Performs content inventory via sitemap crawl and Ahrefs top pages, scores content performance, detects content decay, analyzes topic gaps vs competitors, maps topic clusters, and generates content briefs. Supports Korean content patterns (Naver Blog format, review/후기 content).
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Content audit
+python scripts/content_auditor.py --url https://example.com --json
+
+# Content gap analysis
+python scripts/content_gap_analyzer.py --target https://example.com --competitor https://competitor.com --json
+
+# Generate content brief
+python scripts/content_brief_generator.py --keyword "치과 임플란트 비용" --url https://example.com --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `content_auditor.py` | Content inventory, performance scoring, decay detection | Content inventory with scores and decay flags |
+| `content_gap_analyzer.py` | Topic gap analysis and cluster mapping vs competitors | Missing topics, cluster map, editorial calendar |
+| `content_brief_generator.py` | Generate SEO content briefs with outlines | Brief with outline, keywords, word count targets |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## Content Auditor
+
+```bash
+# Full content audit
+python scripts/content_auditor.py --url https://example.com --json
+
+# Detect decaying content
+python scripts/content_auditor.py --url https://example.com --decay --json
+
+# Filter by content type
+python scripts/content_auditor.py --url https://example.com --type blog --json
+```
+
+**Capabilities**:
+- Content inventory via sitemap crawl + Ahrefs top-pages
+- Performance scoring (traffic, rankings, backlinks)
+- Content decay detection (pages losing traffic over time)
+- Content type classification (blog, product, service, landing, resource)
+- Word count and freshness assessment
+- Korean content format analysis (Naver Blog style, 후기/review content)
+
+## Content Gap Analyzer
+
+```bash
+# Gap analysis vs competitor
+python scripts/content_gap_analyzer.py --target https://example.com --competitor https://comp1.com --json
+
+# With topic cluster mapping
+python scripts/content_gap_analyzer.py --target https://example.com --competitor https://comp1.com --clusters --json
+```
+
+**Capabilities**:
+- Topic gap identification vs competitors
+- Topic cluster mapping (pillar + cluster pages)
+- Content freshness comparison
+- Content volume comparison
+- Editorial calendar generation with priority scoring
+- Korean content opportunity detection
+
+## Content Brief Generator
+
+```bash
+# Generate brief for keyword
+python scripts/content_brief_generator.py --keyword "치과 임플란트 비용" --url https://example.com --json
+
+# With competitor analysis
+python scripts/content_brief_generator.py --keyword "dental implant cost" --url https://example.com --competitors 5 --json
+```
+
+**Capabilities**:
+- Content outline generation with H2/H3 structure
+- Target keyword list (primary + secondary + LSI)
+- Word count recommendation based on top-ranking pages
+- Competitor content analysis (structure, word count, topics covered)
+- Internal linking suggestions
+- Korean content format recommendations
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `site-explorer-top-pages` | Get top performing pages |
+| `site-explorer-pages-by-traffic` | Pages ranked by organic traffic |
+| `site-explorer-organic-keywords` | Keywords per page |
+| `site-explorer-organic-competitors` | Find content competitors |
+| `site-explorer-best-by-external-links` | Best content by links |
+
+## Output Format
+
+```json
+{
+  "url": "https://example.com",
+  "content_inventory": {
+    "total_pages": 150,
+    "by_type": {"blog": 80, "product": 40, "service": 20, "other": 10},
+    "avg_performance_score": 45
+  },
+  "decaying_content": [...],
+  "top_performers": [...],
+  "gaps": [...],
+  "clusters": [...],
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Audited website URL |
+| Category | Select | Content Strategy |
+| Priority | Select | Based on gap severity |
+| Found Date | Date | Audit date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: CONTENT-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is
+- URLs and code remain unchanged
diff --git a/custom-skills/23-seo-content-strategy/code/scripts/base_client.py b/custom-skills/23-seo-content-strategy/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/23-seo-content-strategy/code/scripts/content_auditor.py b/custom-skills/23-seo-content-strategy/code/scripts/content_auditor.py
new file mode 100644
index 0000000..6cb27f9
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/code/scripts/content_auditor.py
@@ -0,0 +1,716 @@
+"""
+Content Auditor - SEO Content Inventory & Performance Analysis
+==============================================================
+Purpose: Build content inventory, score performance, detect decay,
+         classify content types, and analyze Korean content patterns.
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import re
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime, timedelta
+from typing import Any
+from urllib.parse import urlparse
+
+import aiohttp
+import requests
+from bs4 import BeautifulSoup
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class ContentPage:
+    """Single content page with performance metrics."""
+    url: str
+    title: str = ""
+    content_type: str = "other"
+    word_count: int = 0
+    traffic: int = 0
+    keywords_count: int = 0
+    backlinks: int = 0
+    performance_score: float = 0.0
+    last_modified: str = ""
+    is_decaying: bool = False
+    decay_rate: float = 0.0
+    korean_pattern: str = ""
+    topics: list[str] = field(default_factory=list)
+
+
+@dataclass
+class ContentInventory:
+    """Aggregated content inventory summary."""
+    total_pages: int = 0
+    by_type: dict[str, int] = field(default_factory=dict)
+    avg_performance_score: float = 0.0
+    avg_word_count: float = 0.0
+    pages: list[ContentPage] = field(default_factory=list)
+    freshness_distribution: dict[str, int] = field(default_factory=dict)
+
+
+@dataclass
+class ContentAuditResult:
+    """Full content audit result."""
+    url: str
+    timestamp: str = ""
+    content_inventory: ContentInventory = field(default_factory=ContentInventory)
+    top_performers: list[ContentPage] = field(default_factory=list)
+    decaying_content: list[ContentPage] = field(default_factory=list)
+    korean_content_analysis: dict[str, Any] = field(default_factory=dict)
+    recommendations: list[str] = field(default_factory=list)
+    errors: list[str] = field(default_factory=list)
+
+
+# ---------------------------------------------------------------------------
+# URL pattern rules for content type classification
+# ---------------------------------------------------------------------------
+
+CONTENT_TYPE_PATTERNS = {
+    "blog": [
+        r"/blog/", r"/post/", r"/posts/", r"/article/", r"/articles/",
+        r"/news/", r"/magazine/", r"/stories/", r"/insights/",
+        r"/블로그/", r"/소식/", r"/뉴스/",
+    ],
+    "product": [
+        r"/product/", r"/products/", r"/shop/", r"/store/",
+        r"/item/", r"/goods/", r"/catalog/",
+        r"/제품/", r"/상품/", r"/쇼핑/",
+    ],
+    "service": [
+        r"/service/", r"/services/", r"/solutions/", r"/offering/",
+        r"/진료/", r"/서비스/", r"/시술/", r"/치료/",
+    ],
+    "landing": [
+        r"/lp/", r"/landing/", r"/campaign/", r"/promo/",
+        r"/event/", r"/이벤트/", r"/프로모션/",
+    ],
+    "resource": [
+        r"/resource/", r"/resources/", r"/guide/", r"/guides/",
+        r"/whitepaper/", r"/ebook/", r"/download/", r"/faq/",
+        r"/help/", r"/support/", r"/가이드/", r"/자료/",
+    ],
+}
+
+KOREAN_CONTENT_PATTERNS = {
+    "naver_blog_style": [
+        r"후기", r"리뷰", r"체험", r"솔직후기", r"방문후기",
+        r"사용후기", r"이용후기",
+    ],
+    "listicle": [
+        r"추천", r"베스트", r"TOP\s*\d+", r"\d+선", r"\d+가지",
+        r"모음", r"정리", r"비교",
+    ],
+    "how_to": [
+        r"방법", r"하는\s*법", r"하는\s*방법", r"가이드",
+        r"따라하기", r"시작하기", r"알아보기",
+    ],
+    "informational": [
+        r"이란", r"뜻", r"의미", r"차이", r"비교",
+        r"장단점", r"효과", r"부작용", r"비용", r"가격",
+    ],
+}
+
+
+# ---------------------------------------------------------------------------
+# ContentAuditor
+# ---------------------------------------------------------------------------
+
+class ContentAuditor(BaseAsyncClient):
+    """Content auditor using Ahrefs API and sitemap crawling."""
+
+    def __init__(self, max_concurrent: int = 5, requests_per_second: float = 2.0):
+        super().__init__(max_concurrent=max_concurrent, requests_per_second=requests_per_second)
+        self.session: aiohttp.ClientSession | None = None
+
+    async def _ensure_session(self) -> aiohttp.ClientSession:
+        if self.session is None or self.session.closed:
+            timeout = aiohttp.ClientTimeout(total=30)
+            self.session = aiohttp.ClientSession(timeout=timeout)
+        return self.session
+
+    async def close(self) -> None:
+        if self.session and not self.session.closed:
+            await self.session.close()
+
+    # ------------------------------------------------------------------
+    # Ahrefs data retrieval
+    # ------------------------------------------------------------------
+
+    async def get_top_pages(self, url: str, limit: int = 100) -> list[dict]:
+        """
+        Retrieve top pages via Ahrefs site-explorer-top-pages.
+
+        Returns list of dicts with keys: url, traffic, keywords, value, top_keyword.
+        """
+        self.logger.info(f"Fetching top pages from Ahrefs for {url}")
+        target = urlparse(url).netloc or url
+        try:
+            # Ahrefs MCP call: site-explorer-top-pages
+            # In MCP context this would be called by the agent.
+            # Standalone fallback: use REST API if AHREFS_API_KEY is set.
+            api_key = config.get_required("AHREFS_API_KEY") if hasattr(config, "get_required") else None
+            if not api_key:
+                self.logger.warning("AHREFS_API_KEY not set; returning empty top pages")
+                return []
+
+            resp = requests.get(
+                "https://api.ahrefs.com/v3/site-explorer/top-pages",
+                params={"target": target, "limit": limit, "select": "url,traffic,keywords,value,top_keyword"},
+                headers={"Authorization": f"Bearer {api_key}"},
+                timeout=30,
+            )
+            resp.raise_for_status()
+            data = resp.json()
+            pages = data.get("pages", data.get("items", []))
+            self.logger.info(f"Retrieved {len(pages)} top pages")
+            return pages
+        except Exception as exc:
+            self.logger.warning(f"Ahrefs top-pages lookup failed: {exc}")
+            return []
+
+    async def get_pages_by_traffic(self, url: str, limit: int = 100) -> list[dict]:
+        """
+        Retrieve pages sorted by organic traffic via Ahrefs site-explorer-pages-by-traffic.
+
+        Returns list of dicts with keys: url, traffic, keywords, top_keyword.
+        """
+        self.logger.info(f"Fetching pages-by-traffic from Ahrefs for {url}")
+        target = urlparse(url).netloc or url
+        try:
+            api_key = config.get_required("AHREFS_API_KEY") if hasattr(config, "get_required") else None
+            if not api_key:
+                self.logger.warning("AHREFS_API_KEY not set; returning empty traffic pages")
+                return []
+
+            resp = requests.get(
+                "https://api.ahrefs.com/v3/site-explorer/pages-by-traffic",
+                params={"target": target, "limit": limit, "select": "url,traffic,keywords,top_keyword"},
+                headers={"Authorization": f"Bearer {api_key}"},
+                timeout=30,
+            )
+            resp.raise_for_status()
+            data = resp.json()
+            pages = data.get("pages", data.get("items", []))
+            self.logger.info(f"Retrieved {len(pages)} pages by traffic")
+            return pages
+        except Exception as exc:
+            self.logger.warning(f"Ahrefs pages-by-traffic lookup failed: {exc}")
+            return []
+
+    # ------------------------------------------------------------------
+    # Sitemap crawling
+    # ------------------------------------------------------------------
+
+    async def crawl_sitemap(self, url: str) -> list[str]:
+        """Discover URLs from sitemap.xml."""
+        sitemap_urls_to_try = [
+            f"{url.rstrip('/')}/sitemap.xml",
+            f"{url.rstrip('/')}/sitemap_index.xml",
+            f"{url.rstrip('/')}/post-sitemap.xml",
+        ]
+        discovered: list[str] = []
+        session = await self._ensure_session()
+
+        for sitemap_url in sitemap_urls_to_try:
+            try:
+                async with session.get(sitemap_url) as resp:
+                    if resp.status != 200:
+                        continue
+                    text = await resp.text()
+                    soup = BeautifulSoup(text, "lxml-xml")
+
+                    # Sitemap index
+                    sitemaps = soup.find_all("sitemap")
+                    if sitemaps:
+                        for sm in sitemaps:
+                            loc = sm.find("loc")
+                            if loc:
+                                child_urls = await self._parse_sitemap(session, loc.text.strip())
+                                discovered.extend(child_urls)
+                    else:
+                        urls = soup.find_all("url")
+                        for u in urls:
+                            loc = u.find("loc")
+                            if loc:
+                                discovered.append(loc.text.strip())
+
+                    if discovered:
+                        self.logger.info(f"Discovered {len(discovered)} URLs from {sitemap_url}")
+                        break
+            except Exception as exc:
+                self.logger.debug(f"Failed to fetch {sitemap_url}: {exc}")
+
+        return list(set(discovered))
+
+    async def _parse_sitemap(self, session: aiohttp.ClientSession, sitemap_url: str) -> list[str]:
+        """Parse a single sitemap XML and return URLs."""
+        urls: list[str] = []
+        try:
+            async with session.get(sitemap_url) as resp:
+                if resp.status != 200:
+                    return urls
+                text = await resp.text()
+                soup = BeautifulSoup(text, "lxml-xml")
+                for u in soup.find_all("url"):
+                    loc = u.find("loc")
+                    if loc:
+                        urls.append(loc.text.strip())
+        except Exception as exc:
+            self.logger.debug(f"Failed to parse sitemap {sitemap_url}: {exc}")
+        return urls
+
+    # ------------------------------------------------------------------
+    # Content type classification
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def classify_content_type(url: str, title: str = "") -> str:
+        """
+        Classify content type based on URL path patterns and title.
+
+        Returns one of: blog, product, service, landing, resource, other.
+        """
+        combined = f"{url.lower()} {title.lower()}"
+        scores: dict[str, int] = {}
+
+        for ctype, patterns in CONTENT_TYPE_PATTERNS.items():
+            score = 0
+            for pattern in patterns:
+                if re.search(pattern, combined, re.IGNORECASE):
+                    score += 1
+            if score > 0:
+                scores[ctype] = score
+
+        if not scores:
+            return "other"
+        return max(scores, key=scores.get)
+
+    # ------------------------------------------------------------------
+    # Performance scoring
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def score_performance(page: ContentPage) -> float:
+        """
+        Compute composite performance score (0-100) from traffic, keywords, backlinks.
+
+        Weights:
+        - Traffic: 50% (log-scaled, 10k+ traffic = max)
+        - Keywords count: 30% (log-scaled, 500+ = max)
+        - Backlinks: 20% (log-scaled, 100+ = max)
+        """
+        import math
+
+        traffic_score = min(100, (math.log10(max(page.traffic, 1)) / math.log10(10000)) * 100)
+        keywords_score = min(100, (math.log10(max(page.keywords_count, 1)) / math.log10(500)) * 100)
+        backlinks_score = min(100, (math.log10(max(page.backlinks, 1)) / math.log10(100)) * 100)
+
+        composite = (traffic_score * 0.50) + (keywords_score * 0.30) + (backlinks_score * 0.20)
+        return round(min(100, max(0, composite)), 1)
+
+    # ------------------------------------------------------------------
+    # Content decay detection
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def detect_decay(pages: list[ContentPage], threshold: float = -20.0) -> list[ContentPage]:
+        """
+        Flag pages with declining traffic trend.
+
+        Uses a simple heuristic: pages with low performance score relative to
+        their keyword count indicate potential decay. In production, historical
+        traffic data from Ahrefs metrics-history would be used.
+
+        Args:
+            pages: List of content pages with metrics.
+            threshold: Decay rate threshold (percentage decline).
+
+        Returns:
+            List of pages flagged as decaying.
+        """
+        decaying: list[ContentPage] = []
+        for page in pages:
+            # Heuristic: high keyword count but low traffic suggests decay
+            if page.keywords_count > 10 and page.traffic < 50:
+                page.is_decaying = True
+                page.decay_rate = -50.0 if page.traffic == 0 else round(
+                    -((page.keywords_count * 10 - page.traffic) / max(page.keywords_count * 10, 1)) * 100, 1
+                )
+                if page.decay_rate <= threshold:
+                    decaying.append(page)
+            elif page.performance_score < 20 and page.keywords_count > 5:
+                page.is_decaying = True
+                page.decay_rate = round(-max(30, 100 - page.performance_score * 2), 1)
+                if page.decay_rate <= threshold:
+                    decaying.append(page)
+
+        decaying.sort(key=lambda p: p.decay_rate)
+        return decaying
+
+    # ------------------------------------------------------------------
+    # Freshness assessment
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def analyze_freshness(pages: list[ContentPage]) -> dict[str, int]:
+        """
+        Categorize pages by freshness based on last_modified dates.
+
+        Returns distribution: fresh (< 3 months), aging (3-12 months),
+        stale (> 12 months), unknown (no date).
+        """
+        now = datetime.now()
+        distribution = {"fresh": 0, "aging": 0, "stale": 0, "unknown": 0}
+
+        for page in pages:
+            if not page.last_modified:
+                distribution["unknown"] += 1
+                continue
+            try:
+                # Try common date formats
+                for fmt in ("%Y-%m-%dT%H:%M:%S", "%Y-%m-%d", "%Y-%m-%dT%H:%M:%S%z"):
+                    try:
+                        modified = datetime.strptime(
+                            page.last_modified.replace("+00:00", "").replace("Z", ""), fmt.replace("%z", "")
+                        )
+                        break
+                    except ValueError:
+                        continue
+                else:
+                    distribution["unknown"] += 1
+                    continue
+
+                age = now - modified
+                if age < timedelta(days=90):
+                    distribution["fresh"] += 1
+                elif age < timedelta(days=365):
+                    distribution["aging"] += 1
+                else:
+                    distribution["stale"] += 1
+            except Exception:
+                distribution["unknown"] += 1
+
+        return distribution
+
+    # ------------------------------------------------------------------
+    # Korean content pattern identification
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def identify_korean_patterns(pages: list[ContentPage]) -> dict[str, Any]:
+        """
+        Detect Korean content patterns across pages.
+
+        Identifies Naver Blog style review content, listicles,
+        how-to guides, and informational content patterns.
+
+        Returns summary with counts and example URLs per pattern.
+        """
+        results: dict[str, Any] = {
+            "total_korean_content": 0,
+            "patterns": {},
+        }
+
+        for pattern_name, keywords in KOREAN_CONTENT_PATTERNS.items():
+            matches: list[dict[str, str]] = []
+            for page in pages:
+                combined = f"{page.url} {page.title}"
+                for keyword in keywords:
+                    if re.search(keyword, combined, re.IGNORECASE):
+                        matches.append({"url": page.url, "title": page.title, "matched_keyword": keyword})
+                        break
+
+            results["patterns"][pattern_name] = {
+                "count": len(matches),
+                "examples": matches[:5],
+            }
+
+        korean_urls = set()
+        for pattern_data in results["patterns"].values():
+            for example in pattern_data["examples"]:
+                korean_urls.add(example["url"])
+        results["total_korean_content"] = len(korean_urls)
+
+        return results
+
+    # ------------------------------------------------------------------
+    # Orchestration
+    # ------------------------------------------------------------------
+
+    async def audit(
+        self,
+        url: str,
+        detect_decay_flag: bool = False,
+        content_type_filter: str | None = None,
+        limit: int = 200,
+    ) -> ContentAuditResult:
+        """
+        Run full content audit: inventory, scoring, decay, Korean patterns.
+
+        Args:
+            url: Target website URL.
+            detect_decay_flag: Whether to run decay detection.
+            content_type_filter: Filter by content type (blog, product, etc.).
+            limit: Maximum pages to analyze.
+
+        Returns:
+            ContentAuditResult with inventory, top performers, decay, analysis.
+        """
+        result = ContentAuditResult(
+            url=url,
+            timestamp=datetime.now().isoformat(),
+        )
+
+        self.logger.info(f"Starting content audit for {url}")
+
+        # 1. Gather pages from Ahrefs and sitemap
+        top_pages_data, traffic_pages_data, sitemap_urls = await asyncio.gather(
+            self.get_top_pages(url, limit=limit),
+            self.get_pages_by_traffic(url, limit=limit),
+            self.crawl_sitemap(url),
+        )
+
+        # 2. Merge and deduplicate pages
+        page_map: dict[str, ContentPage] = {}
+
+        for item in top_pages_data:
+            page_url = item.get("url", "")
+            if not page_url:
+                continue
+            page_map[page_url] = ContentPage(
+                url=page_url,
+                title=item.get("top_keyword", ""),
+                traffic=int(item.get("traffic", 0)),
+                keywords_count=int(item.get("keywords", 0)),
+                backlinks=int(item.get("value", 0)),
+            )
+
+        for item in traffic_pages_data:
+            page_url = item.get("url", "")
+            if not page_url:
+                continue
+            if page_url in page_map:
+                existing = page_map[page_url]
+                existing.traffic = max(existing.traffic, int(item.get("traffic", 0)))
+                existing.keywords_count = max(existing.keywords_count, int(item.get("keywords", 0)))
+            else:
+                page_map[page_url] = ContentPage(
+                    url=page_url,
+                    title=item.get("top_keyword", ""),
+                    traffic=int(item.get("traffic", 0)),
+                    keywords_count=int(item.get("keywords", 0)),
+                )
+
+        # Add sitemap URLs not already present
+        for s_url in sitemap_urls:
+            if s_url not in page_map:
+                page_map[s_url] = ContentPage(url=s_url)
+
+        # 3. Classify and score
+        all_pages: list[ContentPage] = []
+        for page in page_map.values():
+            page.content_type = self.classify_content_type(page.url, page.title)
+            page.performance_score = self.score_performance(page)
+            all_pages.append(page)
+
+        # 4. Filter by content type if requested
+        if content_type_filter:
+            all_pages = [p for p in all_pages if p.content_type == content_type_filter]
+
+        # 5. Build inventory
+        by_type: dict[str, int] = {}
+        for page in all_pages:
+            by_type[page.content_type] = by_type.get(page.content_type, 0) + 1
+
+        avg_score = (
+            sum(p.performance_score for p in all_pages) / len(all_pages)
+            if all_pages else 0.0
+        )
+        avg_word_count = (
+            sum(p.word_count for p in all_pages) / len(all_pages)
+            if all_pages else 0.0
+        )
+
+        freshness = self.analyze_freshness(all_pages)
+
+        result.content_inventory = ContentInventory(
+            total_pages=len(all_pages),
+            by_type=by_type,
+            avg_performance_score=round(avg_score, 1),
+            avg_word_count=round(avg_word_count, 1),
+            pages=sorted(all_pages, key=lambda p: p.performance_score, reverse=True)[:limit],
+            freshness_distribution=freshness,
+        )
+
+        # 6. Top performers
+        result.top_performers = sorted(all_pages, key=lambda p: p.performance_score, reverse=True)[:20]
+
+        # 7. Decay detection
+        if detect_decay_flag:
+            result.decaying_content = self.detect_decay(all_pages)
+
+        # 8. Korean content analysis
+        result.korean_content_analysis = self.identify_korean_patterns(all_pages)
+
+        # 9. Recommendations
+        result.recommendations = self._generate_recommendations(result)
+
+        self.logger.info(
+            f"Audit complete: {len(all_pages)} pages, "
+            f"{len(result.top_performers)} top performers, "
+            f"{len(result.decaying_content)} decaying"
+        )
+
+        return result
+
+    @staticmethod
+    def _generate_recommendations(result: ContentAuditResult) -> list[str]:
+        """Generate actionable recommendations from audit data."""
+        recs: list[str] = []
+        inv = result.content_inventory
+
+        # Low average score
+        if inv.avg_performance_score < 30:
+            recs.append(
+                "전체 콘텐츠 평균 성과 점수가 낮습니다 ({:.0f}/100). "
+                "상위 콘텐츠 패턴을 분석하여 저성과 페이지를 개선하세요.".format(inv.avg_performance_score)
+            )
+
+        # Stale content
+        stale = inv.freshness_distribution.get("stale", 0)
+        total = inv.total_pages or 1
+        if stale / total > 0.3:
+            recs.append(
+                f"오래된 콘텐츠가 {stale}개 ({stale * 100 // total}%)입니다. "
+                "콘텐츠 업데이트 또는 통합을 고려하세요."
+            )
+
+        # Decaying content
+        if len(result.decaying_content) > 5:
+            recs.append(
+                f"트래픽이 감소하는 콘텐츠가 {len(result.decaying_content)}개 감지되었습니다. "
+                "상위 감소 페이지부터 콘텐츠 리프레시를 진행하세요."
+            )
+
+        # Content type balance
+        blog_count = inv.by_type.get("blog", 0)
+        if blog_count == 0:
+            recs.append(
+                "블로그 콘텐츠가 없습니다. SEO 트래픽 확보를 위해 "
+                "블로그 콘텐츠 전략을 수립하세요."
+            )
+
+        # Korean content opportunities
+        korean = result.korean_content_analysis
+        review_count = korean.get("patterns", {}).get("naver_blog_style", {}).get("count", 0)
+        if review_count == 0:
+            recs.append(
+                "후기/리뷰 콘텐츠가 없습니다. 한국 시장에서 후기 콘텐츠는 "
+                "전환율에 큰 영향을 미치므로 후기 콘텐츠 생성을 권장합니다."
+            )
+
+        if not recs:
+            recs.append("현재 콘텐츠 전략이 양호합니다. 지속적인 모니터링을 권장합니다.")
+
+        return recs
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="SEO Content Auditor - inventory, scoring, and decay detection",
+    )
+    parser.add_argument("--url", required=True, help="Target website URL")
+    parser.add_argument("--decay", action="store_true", help="Enable content decay detection")
+    parser.add_argument("--type", dest="content_type", help="Filter by content type (blog, product, service, landing, resource)")
+    parser.add_argument("--limit", type=int, default=200, help="Maximum pages to analyze (default: 200)")
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", help="Save output to file")
+    return parser
+
+
+def format_text_report(result: ContentAuditResult) -> str:
+    """Format audit result as human-readable text."""
+    lines: list[str] = []
+    lines.append(f"## Content Audit: {result.url}")
+    lines.append(f"**Date**: {result.timestamp[:10]}")
+    lines.append("")
+
+    inv = result.content_inventory
+    lines.append(f"### Content Inventory")
+    lines.append(f"- Total pages: {inv.total_pages}")
+    lines.append(f"- Average performance score: {inv.avg_performance_score}/100")
+    lines.append(f"- Content types: {json.dumps(inv.by_type, ensure_ascii=False)}")
+    lines.append(f"- Freshness: {json.dumps(inv.freshness_distribution, ensure_ascii=False)}")
+    lines.append("")
+
+    lines.append("### Top Performers")
+    for i, page in enumerate(result.top_performers[:10], 1):
+        lines.append(f"  {i}. [{page.performance_score:.0f}] {page.url} (traffic: {page.traffic})")
+    lines.append("")
+
+    if result.decaying_content:
+        lines.append("### Decaying Content")
+        for i, page in enumerate(result.decaying_content[:10], 1):
+            lines.append(f"  {i}. [{page.decay_rate:+.0f}%] {page.url} (traffic: {page.traffic})")
+        lines.append("")
+
+    if result.korean_content_analysis.get("patterns"):
+        lines.append("### Korean Content Patterns")
+        for pattern_name, data in result.korean_content_analysis["patterns"].items():
+            lines.append(f"  - {pattern_name}: {data['count']} pages")
+        lines.append("")
+
+    lines.append("### Recommendations")
+    for i, rec in enumerate(result.recommendations, 1):
+        lines.append(f"  {i}. {rec}")
+
+    return "\n".join(lines)
+
+
+async def main() -> None:
+    parser = build_parser()
+    args = parser.parse_args()
+
+    auditor = ContentAuditor()
+    try:
+        result = await auditor.audit(
+            url=args.url,
+            detect_decay_flag=args.decay,
+            content_type_filter=args.content_type,
+            limit=args.limit,
+        )
+
+        if args.json:
+            output = json.dumps(asdict(result), ensure_ascii=False, indent=2, default=str)
+        else:
+            output = format_text_report(result)
+
+        if args.output:
+            with open(args.output, "w", encoding="utf-8") as f:
+                f.write(output)
+            logger.info(f"Output saved to {args.output}")
+        else:
+            print(output)
+
+    finally:
+        await auditor.close()
+        auditor.print_stats()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/23-seo-content-strategy/code/scripts/content_brief_generator.py b/custom-skills/23-seo-content-strategy/code/scripts/content_brief_generator.py
new file mode 100644
index 0000000..f4dda60
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/code/scripts/content_brief_generator.py
@@ -0,0 +1,738 @@
+"""
+Content Brief Generator - SEO Content Brief Creation
+=====================================================
+Purpose: Generate detailed SEO content briefs with outlines,
+         keyword lists, word count targets, and internal linking suggestions.
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import math
+import re
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from typing import Any
+from urllib.parse import urlparse
+
+import aiohttp
+import requests
+from bs4 import BeautifulSoup
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class OutlineSection:
+    """A single heading section in the content outline."""
+    heading: str
+    level: int = 2  # H2 or H3
+    talking_points: list[str] = field(default_factory=list)
+    target_words: int = 200
+    keywords_to_include: list[str] = field(default_factory=list)
+
+
+@dataclass
+class CompetitorPageAnalysis:
+    """Analysis of a single competitor page for the target keyword."""
+    url: str
+    title: str = ""
+    word_count: int = 0
+    headings: list[dict[str, str]] = field(default_factory=list)
+    topics_covered: list[str] = field(default_factory=list)
+    content_type: str = ""
+    has_images: bool = False
+    has_video: bool = False
+    has_faq: bool = False
+    has_table: bool = False
+
+
+@dataclass
+class ContentBrief:
+    """Complete SEO content brief."""
+    primary_keyword: str
+    secondary_keywords: list[str] = field(default_factory=list)
+    lsi_keywords: list[str] = field(default_factory=list)
+    target_word_count: int = 1500
+    word_count_range: tuple[int, int] = (1200, 1800)
+    suggested_title: str = ""
+    meta_description: str = ""
+    outline: list[OutlineSection] = field(default_factory=list)
+    competitor_analysis: list[CompetitorPageAnalysis] = field(default_factory=list)
+    internal_links: list[dict[str, str]] = field(default_factory=list)
+    content_format: str = "blog"
+    korean_format_recommendations: list[str] = field(default_factory=list)
+    search_intent: str = "informational"
+    notes: list[str] = field(default_factory=list)
+    timestamp: str = ""
+
+
+# ---------------------------------------------------------------------------
+# Search intent patterns
+# ---------------------------------------------------------------------------
+
+INTENT_PATTERNS = {
+    "transactional": [
+        r"buy", r"purchase", r"price", r"cost", r"order", r"shop",
+        r"구매", r"주문", r"가격", r"비용", r"할인", r"쿠폰",
+    ],
+    "navigational": [
+        r"login", r"sign in", r"official", r"website",
+        r"로그인", r"공식", r"홈페이지",
+    ],
+    "commercial": [
+        r"best", r"top", r"review", r"compare", r"vs",
+        r"추천", r"비교", r"후기", r"리뷰", r"순위",
+    ],
+    "informational": [
+        r"what", r"how", r"why", r"guide", r"tutorial",
+        r"이란", r"방법", r"가이드", r"효과", r"원인",
+    ],
+}
+
+# ---------------------------------------------------------------------------
+# Korean content format recommendations
+# ---------------------------------------------------------------------------
+
+KOREAN_FORMAT_TIPS = {
+    "transactional": [
+        "가격 비교표를 포함하세요 (경쟁사 가격 대비)",
+        "실제 비용 사례를 3개 이상 제시하세요",
+        "결제 방법 및 할인 정보를 명확히 안내하세요",
+        "CTA(행동 유도) 버튼을 여러 위치에 배치하세요",
+    ],
+    "commercial": [
+        "네이버 블로그 스타일의 솔직한 후기 톤을 사용하세요",
+        "장단점을 균형 있게 비교하세요",
+        "실제 사용 사진 또는 전후 비교 이미지를 포함하세요",
+        "별점 또는 점수 평가 체계를 추가하세요",
+        "FAQ 섹션을 포함하세요 (네이버 검색 노출에 유리)",
+    ],
+    "informational": [
+        "핵심 정보를 글 상단에 요약하세요 (두괄식 구성)",
+        "전문 용어는 쉬운 설명을 병기하세요",
+        "인포그래픽 또는 도표를 활용하세요",
+        "관련 콘텐츠 내부 링크를 3-5개 포함하세요",
+        "전문가 인용 또는 출처를 명시하세요 (E-E-A-T 강화)",
+    ],
+    "navigational": [
+        "공식 정보와 연락처를 최상단에 배치하세요",
+        "지도 임베드를 포함하세요 (네이버 지도/구글 맵)",
+        "영업시간, 주소, 전화번호를 명확히 표시하세요",
+    ],
+}
+
+
+# ---------------------------------------------------------------------------
+# ContentBriefGenerator
+# ---------------------------------------------------------------------------
+
+class ContentBriefGenerator(BaseAsyncClient):
+    """Generate comprehensive SEO content briefs."""
+
+    def __init__(self, max_concurrent: int = 5, requests_per_second: float = 2.0):
+        super().__init__(max_concurrent=max_concurrent, requests_per_second=requests_per_second)
+        self.session: aiohttp.ClientSession | None = None
+
+    async def _ensure_session(self) -> aiohttp.ClientSession:
+        if self.session is None or self.session.closed:
+            timeout = aiohttp.ClientTimeout(total=30)
+            headers = {
+                "User-Agent": "Mozilla/5.0 (compatible; SEOContentBrief/1.0)",
+            }
+            self.session = aiohttp.ClientSession(timeout=timeout, headers=headers)
+        return self.session
+
+    async def close(self) -> None:
+        if self.session and not self.session.closed:
+            await self.session.close()
+
+    # ------------------------------------------------------------------
+    # Analyze top ranking results
+    # ------------------------------------------------------------------
+
+    async def analyze_top_results(
+        self,
+        keyword: str,
+        site_url: str | None = None,
+        num_competitors: int = 5,
+    ) -> list[CompetitorPageAnalysis]:
+        """
+        Analyze top ranking pages for a keyword using Ahrefs SERP data.
+
+        Falls back to fetching pages directly if Ahrefs data is unavailable.
+        """
+        self.logger.info(f"Analyzing top results for: {keyword}")
+        results: list[CompetitorPageAnalysis] = []
+
+        # Try Ahrefs organic keywords to find ranking pages
+        try:
+            api_key = config.get_required("AHREFS_API_KEY") if hasattr(config, "get_required") else None
+            if api_key:
+                resp = requests.get(
+                    "https://api.ahrefs.com/v3/serp-overview",
+                    params={"keyword": keyword, "select": "url,title,position,traffic"},
+                    headers={"Authorization": f"Bearer {api_key}"},
+                    timeout=30,
+                )
+                if resp.status_code == 200:
+                    data = resp.json()
+                    serp_items = data.get("positions", data.get("items", []))[:num_competitors]
+                    for item in serp_items:
+                        analysis = CompetitorPageAnalysis(
+                            url=item.get("url", ""),
+                            title=item.get("title", ""),
+                        )
+                        results.append(analysis)
+        except Exception as exc:
+            self.logger.warning(f"Ahrefs SERP lookup failed: {exc}")
+
+        # Fetch and analyze each page
+        session = await self._ensure_session()
+        for analysis in results[:num_competitors]:
+            if not analysis.url:
+                continue
+            try:
+                async with session.get(analysis.url) as resp:
+                    if resp.status != 200:
+                        continue
+                    html = await resp.text()
+                    self._analyze_page_content(analysis, html)
+            except Exception as exc:
+                self.logger.debug(f"Failed to fetch {analysis.url}: {exc}")
+
+        self.logger.info(f"Analyzed {len(results)} competitor pages")
+        return results
+
+    @staticmethod
+    def _analyze_page_content(analysis: CompetitorPageAnalysis, html: str) -> None:
+        """Parse HTML and extract content metrics."""
+        soup = BeautifulSoup(html, "html.parser")
+
+        # Title
+        title_tag = soup.find("title")
+        if title_tag and not analysis.title:
+            analysis.title = title_tag.get_text(strip=True)
+
+        # Word count (visible text only)
+        for tag in soup(["script", "style", "nav", "header", "footer"]):
+            tag.decompose()
+        visible_text = soup.get_text(separator=" ", strip=True)
+        analysis.word_count = len(visible_text.split())
+
+        # Headings
+        headings: list[dict[str, str]] = []
+        for level in range(1, 7):
+            for h in soup.find_all(f"h{level}"):
+                text = h.get_text(strip=True)
+                if text:
+                    headings.append({"level": f"H{level}", "text": text})
+        analysis.headings = headings
+
+        # Content features
+        analysis.has_images = len(soup.find_all("img")) > 2
+        analysis.has_video = bool(soup.find("video") or soup.find("iframe", src=re.compile(r"youtube|vimeo")))
+        analysis.has_faq = bool(
+            soup.find(string=re.compile(r"FAQ|자주\s*묻는\s*질문|Q\s*&\s*A", re.IGNORECASE))
+            or soup.find("script", type="application/ld+json", string=re.compile(r"FAQPage"))
+        )
+        analysis.has_table = bool(soup.find("table"))
+
+        # Topics covered (from H2 headings)
+        analysis.topics_covered = [
+            h["text"] for h in headings if h["level"] == "H2"
+        ][:15]
+
+    # ------------------------------------------------------------------
+    # Extract content outline
+    # ------------------------------------------------------------------
+
+    def extract_outline(
+        self,
+        keyword: str,
+        top_results: list[CompetitorPageAnalysis],
+    ) -> list[OutlineSection]:
+        """
+        Build recommended H2/H3 outline by aggregating competitor headings.
+
+        Identifies common topics across top-ranking pages and structures
+        them into a logical outline.
+        """
+        # Collect all H2 headings
+        h2_topics: dict[str, int] = {}
+        h3_by_h2: dict[str, list[str]] = {}
+
+        for result in top_results:
+            current_h2 = ""
+            for heading in result.headings:
+                text = heading["text"].strip()
+                if heading["level"] == "H2":
+                    current_h2 = text
+                    h2_topics[text] = h2_topics.get(text, 0) + 1
+                elif heading["level"] == "H3" and current_h2:
+                    if current_h2 not in h3_by_h2:
+                        h3_by_h2[current_h2] = []
+                    h3_by_h2[current_h2].append(text)
+
+        # Sort H2s by frequency (most common topics first)
+        sorted_h2s = sorted(h2_topics.items(), key=lambda x: x[1], reverse=True)
+
+        # Build outline
+        outline: list[OutlineSection] = []
+        target_word_count = self.calculate_word_count(top_results)
+        words_per_section = target_word_count // max(len(sorted_h2s), 5)
+
+        for h2_text, frequency in sorted_h2s[:8]:
+            section = OutlineSection(
+                heading=h2_text,
+                level=2,
+                target_words=words_per_section,
+                talking_points=[],
+            )
+
+            # Add H3 subtopics
+            if h2_text in h3_by_h2:
+                unique_h3s = list(dict.fromkeys(h3_by_h2[h2_text]))[:5]
+                for h3_text in unique_h3s:
+                    subsection = OutlineSection(
+                        heading=h3_text,
+                        level=3,
+                        target_words=words_per_section // 3,
+                    )
+                    section.talking_points.append(h3_text)
+
+            outline.append(section)
+
+        # Ensure FAQ section if common
+        faq_count = sum(1 for r in top_results if r.has_faq)
+        if faq_count >= 2 and not any("FAQ" in s.heading or "질문" in s.heading for s in outline):
+            outline.append(OutlineSection(
+                heading="자주 묻는 질문 (FAQ)",
+                level=2,
+                target_words=300,
+                talking_points=[
+                    f"{keyword} 관련 자주 묻는 질문 5-7개",
+                    "Schema markup (FAQPage) 적용 권장",
+                ],
+            ))
+
+        return outline
+
+    # ------------------------------------------------------------------
+    # Keyword suggestions
+    # ------------------------------------------------------------------
+
+    async def suggest_keywords(self, primary_keyword: str) -> dict[str, list[str]]:
+        """
+        Generate primary, secondary, and LSI keyword suggestions.
+
+        Uses Ahrefs related keywords and matching terms data.
+        """
+        self.logger.info(f"Generating keyword suggestions for: {primary_keyword}")
+        result = {
+            "primary": [primary_keyword],
+            "secondary": [],
+            "lsi": [],
+        }
+
+        try:
+            api_key = config.get_required("AHREFS_API_KEY") if hasattr(config, "get_required") else None
+            if not api_key:
+                self.logger.warning("AHREFS_API_KEY not set; returning basic keywords only")
+                return result
+
+            # Matching terms
+            resp = requests.get(
+                "https://api.ahrefs.com/v3/keywords-explorer/matching-terms",
+                params={"keyword": primary_keyword, "limit": 20, "select": "keyword,volume,difficulty"},
+                headers={"Authorization": f"Bearer {api_key}"},
+                timeout=30,
+            )
+            if resp.status_code == 200:
+                data = resp.json()
+                terms = data.get("keywords", data.get("items", []))
+                for term in terms:
+                    kw = term.get("keyword", "")
+                    if kw and kw.lower() != primary_keyword.lower():
+                        result["secondary"].append(kw)
+
+            # Related terms (LSI)
+            resp2 = requests.get(
+                "https://api.ahrefs.com/v3/keywords-explorer/related-terms",
+                params={"keyword": primary_keyword, "limit": 15, "select": "keyword,volume"},
+                headers={"Authorization": f"Bearer {api_key}"},
+                timeout=30,
+            )
+            if resp2.status_code == 200:
+                data2 = resp2.json()
+                related = data2.get("keywords", data2.get("items", []))
+                for term in related:
+                    kw = term.get("keyword", "")
+                    if kw and kw not in result["secondary"]:
+                        result["lsi"].append(kw)
+
+        except Exception as exc:
+            self.logger.warning(f"Keyword suggestion lookup failed: {exc}")
+
+        return result
+
+    # ------------------------------------------------------------------
+    # Word count calculation
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def calculate_word_count(top_results: list[CompetitorPageAnalysis]) -> int:
+        """
+        Calculate target word count based on top 5 ranking pages.
+
+        Returns the average word count of top 5 with +/- 20% range.
+        """
+        word_counts = [r.word_count for r in top_results[:5] if r.word_count > 100]
+
+        if not word_counts:
+            return 1500  # Default fallback
+
+        avg = sum(word_counts) / len(word_counts)
+        # Round to nearest 100
+        target = round(avg / 100) * 100
+        return max(800, min(5000, target))
+
+    # ------------------------------------------------------------------
+    # Internal linking suggestions
+    # ------------------------------------------------------------------
+
+    async def suggest_internal_links(
+        self,
+        keyword: str,
+        site_url: str,
+    ) -> list[dict[str, str]]:
+        """
+        Find related existing pages on the site for internal linking.
+
+        Uses Ahrefs organic keywords to find pages ranking for related terms.
+        """
+        self.logger.info(f"Finding internal link opportunities for {keyword} on {site_url}")
+        links: list[dict[str, str]] = []
+        target = urlparse(site_url).netloc or site_url
+
+        try:
+            api_key = config.get_required("AHREFS_API_KEY") if hasattr(config, "get_required") else None
+            if not api_key:
+                return links
+
+            resp = requests.get(
+                "https://api.ahrefs.com/v3/site-explorer/organic-keywords",
+                params={
+                    "target": target,
+                    "limit": 50,
+                    "select": "keyword,url,position,traffic",
+                },
+                headers={"Authorization": f"Bearer {api_key}"},
+                timeout=30,
+            )
+            if resp.status_code != 200:
+                return links
+
+            data = resp.json()
+            keywords_data = data.get("keywords", data.get("items", []))
+
+            # Find pages ranking for related keywords
+            keyword_lower = keyword.lower()
+            keyword_words = set(keyword_lower.split())
+
+            seen_urls: set[str] = set()
+            for item in keywords_data:
+                kw = item.get("keyword", "").lower()
+                url = item.get("url", "")
+
+                if not url or url in seen_urls:
+                    continue
+
+                # Check keyword relevance
+                kw_words = set(kw.split())
+                overlap = keyword_words & kw_words
+                if overlap and kw != keyword_lower:
+                    links.append({
+                        "url": url,
+                        "anchor_text": kw,
+                        "relevance": f"{len(overlap)}/{len(keyword_words)} word overlap",
+                        "current_traffic": str(item.get("traffic", 0)),
+                    })
+                    seen_urls.add(url)
+
+            links.sort(key=lambda l: int(l.get("current_traffic", "0")), reverse=True)
+
+        except Exception as exc:
+            self.logger.warning(f"Internal link suggestion failed: {exc}")
+
+        return links[:10]
+
+    # ------------------------------------------------------------------
+    # Search intent detection
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def detect_search_intent(keyword: str) -> str:
+        """Classify keyword search intent."""
+        keyword_lower = keyword.lower()
+        scores: dict[str, int] = {}
+
+        for intent, patterns in INTENT_PATTERNS.items():
+            score = sum(1 for p in patterns if re.search(p, keyword_lower, re.IGNORECASE))
+            if score > 0:
+                scores[intent] = score
+
+        if not scores:
+            return "informational"
+        return max(scores, key=scores.get)
+
+    # ------------------------------------------------------------------
+    # Orchestration
+    # ------------------------------------------------------------------
+
+    async def generate(
+        self,
+        keyword: str,
+        site_url: str,
+        num_competitors: int = 5,
+    ) -> ContentBrief:
+        """
+        Generate a comprehensive SEO content brief.
+
+        Args:
+            keyword: Primary target keyword.
+            site_url: Target website URL.
+            num_competitors: Number of competitor pages to analyze.
+
+        Returns:
+            ContentBrief with outline, keywords, and recommendations.
+        """
+        self.logger.info(f"Generating content brief for: {keyword}")
+
+        # Detect search intent
+        intent = self.detect_search_intent(keyword)
+
+        # Run analyses in parallel
+        top_results_task = self.analyze_top_results(keyword, site_url, num_competitors)
+        keywords_task = self.suggest_keywords(keyword)
+        internal_links_task = self.suggest_internal_links(keyword, site_url)
+
+        top_results, keyword_data, internal_links = await asyncio.gather(
+            top_results_task, keywords_task, internal_links_task,
+        )
+
+        # Calculate word count target
+        target_word_count = self.calculate_word_count(top_results)
+        word_count_min = int(target_word_count * 0.8)
+        word_count_max = int(target_word_count * 1.2)
+
+        # Build outline
+        outline = self.extract_outline(keyword, top_results)
+
+        # Generate title suggestion
+        suggested_title = self._generate_title(keyword, intent)
+
+        # Generate meta description
+        meta_description = self._generate_meta_description(keyword, intent)
+
+        # Korean format recommendations
+        korean_tips = KOREAN_FORMAT_TIPS.get(intent, KOREAN_FORMAT_TIPS["informational"])
+
+        brief = ContentBrief(
+            primary_keyword=keyword,
+            secondary_keywords=keyword_data.get("secondary", [])[:10],
+            lsi_keywords=keyword_data.get("lsi", [])[:10],
+            target_word_count=target_word_count,
+            word_count_range=(word_count_min, word_count_max),
+            suggested_title=suggested_title,
+            meta_description=meta_description,
+            outline=outline,
+            competitor_analysis=top_results,
+            internal_links=internal_links,
+            content_format=self._suggest_format(intent, top_results),
+            korean_format_recommendations=korean_tips,
+            search_intent=intent,
+            timestamp=datetime.now().isoformat(),
+        )
+
+        self.logger.info(
+            f"Brief generated: {len(outline)} sections, "
+            f"{target_word_count} target words, "
+            f"{len(keyword_data.get('secondary', []))} secondary keywords"
+        )
+
+        return brief
+
+    @staticmethod
+    def _generate_title(keyword: str, intent: str) -> str:
+        """Generate a suggested title based on keyword and intent."""
+        templates = {
+            "informational": "{keyword} - 완벽 가이드 (2025년 최신)",
+            "commercial": "{keyword} 추천 TOP 10 비교 (전문가 리뷰)",
+            "transactional": "{keyword} 가격 비교 및 구매 가이드",
+            "navigational": "{keyword} - 공식 안내",
+        }
+        template = templates.get(intent, templates["informational"])
+        return template.format(keyword=keyword)
+
+    @staticmethod
+    def _generate_meta_description(keyword: str, intent: str) -> str:
+        """Generate a suggested meta description."""
+        templates = {
+            "informational": (
+                f"{keyword}에 대해 알아야 할 모든 것을 정리했습니다. "
+                "전문가가 알려주는 핵심 정보와 실용적인 가이드를 확인하세요."
+            ),
+            "commercial": (
+                f"{keyword} 비교 분석! 장단점, 가격, 실제 후기를 "
+                "한눈에 비교하고 최적의 선택을 하세요."
+            ),
+            "transactional": (
+                f"{keyword} 최저가 비교 및 구매 방법을 안내합니다. "
+                "합리적인 가격으로 구매하는 팁을 확인하세요."
+            ),
+            "navigational": (
+                f"{keyword} 공식 정보 및 이용 안내. "
+                "정확한 정보를 빠르게 확인하세요."
+            ),
+        }
+        return templates.get(intent, templates["informational"])
+
+    @staticmethod
+    def _suggest_format(intent: str, results: list[CompetitorPageAnalysis]) -> str:
+        """Suggest content format based on intent and competitor analysis."""
+        if intent == "commercial":
+            return "listicle"
+        if intent == "informational":
+            return "guide"
+        if intent == "transactional":
+            return "landing"
+
+        # Check competitor patterns
+        avg_word_count = (
+            sum(r.word_count for r in results) / len(results) if results else 0
+        )
+        if avg_word_count > 3000:
+            return "comprehensive_guide"
+        return "blog"
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="SEO Content Brief Generator",
+    )
+    parser.add_argument("--keyword", required=True, help="Primary target keyword")
+    parser.add_argument("--url", required=True, help="Target website URL")
+    parser.add_argument("--competitors", type=int, default=5, help="Number of competitor pages to analyze (default: 5)")
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", help="Save output to file")
+    return parser
+
+
+def format_text_report(brief: ContentBrief) -> str:
+    """Format content brief as human-readable text."""
+    lines: list[str] = []
+    lines.append(f"## Content Brief: {brief.primary_keyword}")
+    lines.append(f"**Date**: {brief.timestamp[:10]}")
+    lines.append(f"**Search Intent**: {brief.search_intent}")
+    lines.append(f"**Content Format**: {brief.content_format}")
+    lines.append("")
+
+    lines.append("### Target Metrics")
+    lines.append(f"- Word count: {brief.target_word_count} ({brief.word_count_range[0]}-{brief.word_count_range[1]})")
+    lines.append(f"- Suggested title: {brief.suggested_title}")
+    lines.append(f"- Meta description: {brief.meta_description}")
+    lines.append("")
+
+    lines.append("### Keywords")
+    lines.append(f"- **Primary**: {brief.primary_keyword}")
+    if brief.secondary_keywords:
+        lines.append(f"- **Secondary**: {', '.join(brief.secondary_keywords[:8])}")
+    if brief.lsi_keywords:
+        lines.append(f"- **LSI**: {', '.join(brief.lsi_keywords[:8])}")
+    lines.append("")
+
+    lines.append("### Content Outline")
+    for section in brief.outline:
+        prefix = "##" if section.level == 2 else "###"
+        lines.append(f"  {prefix} {section.heading} (~{section.target_words}w)")
+        for point in section.talking_points:
+            lines.append(f"    - {point}")
+    lines.append("")
+
+    if brief.competitor_analysis:
+        lines.append(f"### Competitor Analysis ({len(brief.competitor_analysis)} pages)")
+        for comp in brief.competitor_analysis:
+            lines.append(f"  - **{comp.title or comp.url}**")
+            lines.append(f"    Word count: {comp.word_count} | Headings: {len(comp.headings)}")
+            features = []
+            if comp.has_images:
+                features.append("images")
+            if comp.has_video:
+                features.append("video")
+            if comp.has_faq:
+                features.append("FAQ")
+            if comp.has_table:
+                features.append("table")
+            if features:
+                lines.append(f"    Features: {', '.join(features)}")
+        lines.append("")
+
+    if brief.internal_links:
+        lines.append(f"### Internal Linking Suggestions ({len(brief.internal_links)})")
+        for link in brief.internal_links[:7]:
+            lines.append(f"  - [{link['anchor_text']}]({link['url']})")
+        lines.append("")
+
+    if brief.korean_format_recommendations:
+        lines.append("### Korean Content Format Recommendations")
+        for tip in brief.korean_format_recommendations:
+            lines.append(f"  - {tip}")
+
+    return "\n".join(lines)
+
+
+async def main() -> None:
+    parser = build_parser()
+    args = parser.parse_args()
+
+    generator = ContentBriefGenerator()
+    try:
+        brief = await generator.generate(
+            keyword=args.keyword,
+            site_url=args.url,
+            num_competitors=args.competitors,
+        )
+
+        if args.json:
+            output = json.dumps(asdict(brief), ensure_ascii=False, indent=2, default=str)
+        else:
+            output = format_text_report(brief)
+
+        if args.output:
+            with open(args.output, "w", encoding="utf-8") as f:
+                f.write(output)
+            logger.info(f"Output saved to {args.output}")
+        else:
+            print(output)
+
+    finally:
+        await generator.close()
+        generator.print_stats()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/23-seo-content-strategy/code/scripts/content_gap_analyzer.py b/custom-skills/23-seo-content-strategy/code/scripts/content_gap_analyzer.py
new file mode 100644
index 0000000..cdaf3a4
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/code/scripts/content_gap_analyzer.py
@@ -0,0 +1,694 @@
+"""
+Content Gap Analyzer - Topic Gap Detection & Cluster Mapping
+=============================================================
+Purpose: Identify content gaps vs competitors, build topic clusters,
+         and generate prioritized editorial calendars.
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import math
+import re
+import sys
+from collections import defaultdict
+from dataclasses import asdict, dataclass, field
+from datetime import datetime, timedelta
+from typing import Any
+from urllib.parse import urlparse
+
+import requests
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.cluster import AgglomerativeClustering
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class TopicGap:
+    """A topic present in competitors but missing from target."""
+    topic: str
+    competitor_urls: list[str] = field(default_factory=list)
+    competitor_keywords: list[str] = field(default_factory=list)
+    estimated_traffic: int = 0
+    priority_score: float = 0.0
+    difficulty: str = "medium"
+    content_type_suggestion: str = "blog"
+
+
+@dataclass
+class TopicCluster:
+    """Topic cluster with pillar and supporting cluster pages."""
+    pillar_topic: str
+    pillar_keyword: str = ""
+    cluster_topics: list[str] = field(default_factory=list)
+    cluster_keywords: list[str] = field(default_factory=list)
+    total_volume: int = 0
+    coverage_score: float = 0.0
+
+
+@dataclass
+class CalendarEntry:
+    """Prioritized editorial calendar entry."""
+    topic: str
+    priority: str = "medium"
+    target_date: str = ""
+    content_type: str = "blog"
+    target_word_count: int = 1500
+    primary_keyword: str = ""
+    estimated_traffic: int = 0
+    cluster_name: str = ""
+    notes: str = ""
+
+
+@dataclass
+class ContentGapResult:
+    """Full content gap analysis result."""
+    target_url: str
+    competitor_urls: list[str] = field(default_factory=list)
+    timestamp: str = ""
+    target_topics_count: int = 0
+    competitor_topics_count: int = 0
+    gaps: list[TopicGap] = field(default_factory=list)
+    clusters: list[TopicCluster] = field(default_factory=list)
+    calendar: list[CalendarEntry] = field(default_factory=list)
+    content_volume_comparison: dict[str, int] = field(default_factory=dict)
+    korean_opportunities: list[dict[str, Any]] = field(default_factory=dict)
+    recommendations: list[str] = field(default_factory=list)
+    errors: list[str] = field(default_factory=list)
+
+
+# ---------------------------------------------------------------------------
+# Korean opportunity patterns
+# ---------------------------------------------------------------------------
+
+KOREAN_OPPORTUNITY_PATTERNS = [
+    {"pattern": r"후기|리뷰", "label": "review_content", "description": "후기/리뷰 콘텐츠"},
+    {"pattern": r"비용|가격|견적", "label": "pricing_content", "description": "비용/가격 정보 콘텐츠"},
+    {"pattern": r"비교|차이", "label": "comparison_content", "description": "비교 콘텐츠"},
+    {"pattern": r"추천|베스트|TOP", "label": "recommendation_content", "description": "추천/리스트 콘텐츠"},
+    {"pattern": r"방법|하는\s*법|가이드", "label": "how_to_content", "description": "가이드/방법 콘텐츠"},
+    {"pattern": r"부작용|주의|위험", "label": "safety_content", "description": "안전/부작용 정보"},
+    {"pattern": r"효과|결과|전후", "label": "results_content", "description": "효과/결과 콘텐츠"},
+]
+
+
+# ---------------------------------------------------------------------------
+# ContentGapAnalyzer
+# ---------------------------------------------------------------------------
+
+class ContentGapAnalyzer(BaseAsyncClient):
+    """Analyze content gaps between target and competitor sites."""
+
+    def __init__(self, max_concurrent: int = 5, requests_per_second: float = 2.0):
+        super().__init__(max_concurrent=max_concurrent, requests_per_second=requests_per_second)
+
+    # ------------------------------------------------------------------
+    # Ahrefs data retrieval
+    # ------------------------------------------------------------------
+
+    async def get_competitor_topics(self, competitor_url: str, limit: int = 100) -> list[dict]:
+        """
+        Get top pages and keywords for a competitor via Ahrefs.
+
+        Returns list of dicts: url, traffic, keywords, top_keyword, title.
+        """
+        self.logger.info(f"Fetching competitor topics for {competitor_url}")
+        target = urlparse(competitor_url).netloc or competitor_url
+
+        try:
+            api_key = config.get_required("AHREFS_API_KEY") if hasattr(config, "get_required") else None
+            if not api_key:
+                self.logger.warning("AHREFS_API_KEY not set; returning empty competitor topics")
+                return []
+
+            resp = requests.get(
+                "https://api.ahrefs.com/v3/site-explorer/top-pages",
+                params={
+                    "target": target,
+                    "limit": limit,
+                    "select": "url,traffic,keywords,value,top_keyword",
+                },
+                headers={"Authorization": f"Bearer {api_key}"},
+                timeout=30,
+            )
+            resp.raise_for_status()
+            data = resp.json()
+            pages = data.get("pages", data.get("items", []))
+            self.logger.info(f"Retrieved {len(pages)} competitor topics from {competitor_url}")
+            return pages
+        except Exception as exc:
+            self.logger.warning(f"Failed to get competitor topics for {competitor_url}: {exc}")
+            return []
+
+    async def get_target_keywords(self, target_url: str, limit: int = 200) -> set[str]:
+        """Get the set of keywords the target site already ranks for."""
+        self.logger.info(f"Fetching target keywords for {target_url}")
+        target = urlparse(target_url).netloc or target_url
+
+        try:
+            api_key = config.get_required("AHREFS_API_KEY") if hasattr(config, "get_required") else None
+            if not api_key:
+                return set()
+
+            resp = requests.get(
+                "https://api.ahrefs.com/v3/site-explorer/organic-keywords",
+                params={"target": target, "limit": limit, "select": "keyword,position,traffic"},
+                headers={"Authorization": f"Bearer {api_key}"},
+                timeout=30,
+            )
+            resp.raise_for_status()
+            data = resp.json()
+            keywords = data.get("keywords", data.get("items", []))
+            return {kw.get("keyword", "").lower() for kw in keywords if kw.get("keyword")}
+        except Exception as exc:
+            self.logger.warning(f"Failed to get target keywords: {exc}")
+            return set()
+
+    async def get_organic_competitors(self, target_url: str, limit: int = 10) -> list[str]:
+        """Discover organic competitors via Ahrefs."""
+        self.logger.info(f"Discovering organic competitors for {target_url}")
+        target = urlparse(target_url).netloc or target_url
+
+        try:
+            api_key = config.get_required("AHREFS_API_KEY") if hasattr(config, "get_required") else None
+            if not api_key:
+                return []
+
+            resp = requests.get(
+                "https://api.ahrefs.com/v3/site-explorer/organic-competitors",
+                params={"target": target, "limit": limit},
+                headers={"Authorization": f"Bearer {api_key}"},
+                timeout=30,
+            )
+            resp.raise_for_status()
+            data = resp.json()
+            competitors = data.get("competitors", data.get("items", []))
+            return [c.get("domain", "") for c in competitors if c.get("domain")]
+        except Exception as exc:
+            self.logger.warning(f"Failed to discover competitors: {exc}")
+            return []
+
+    # ------------------------------------------------------------------
+    # Gap analysis
+    # ------------------------------------------------------------------
+
+    async def find_topic_gaps(
+        self,
+        target_url: str,
+        competitor_urls: list[str],
+    ) -> tuple[list[TopicGap], set[str], dict[str, int]]:
+        """
+        Identify topics covered by competitors but missing from target.
+
+        Returns:
+            - List of TopicGap objects.
+            - Set of target keywords (for reference).
+            - Content volume comparison dict.
+        """
+        # Gather target keywords
+        target_keywords = await self.get_target_keywords(target_url)
+
+        # Gather competitor data in parallel
+        competitor_tasks = [self.get_competitor_topics(c_url) for c_url in competitor_urls]
+        competitor_results = await asyncio.gather(*competitor_tasks, return_exceptions=True)
+
+        # Build competitor topic map
+        competitor_topic_map: dict[str, TopicGap] = {}
+        content_volume: dict[str, int] = {target_url: len(target_keywords)}
+
+        for c_url, c_result in zip(competitor_urls, competitor_results):
+            if isinstance(c_result, Exception):
+                self.logger.warning(f"Error fetching {c_url}: {c_result}")
+                continue
+
+            pages = c_result if isinstance(c_result, list) else []
+            content_volume[c_url] = len(pages)
+
+            for page in pages:
+                top_keyword = page.get("top_keyword", "").strip().lower()
+                if not top_keyword:
+                    continue
+
+                # Skip if target already covers this keyword
+                if top_keyword in target_keywords:
+                    continue
+
+                # Check for fuzzy matches (keyword contained in target set)
+                is_covered = any(
+                    top_keyword in tk or tk in top_keyword
+                    for tk in target_keywords
+                    if len(tk) > 3
+                )
+                if is_covered:
+                    continue
+
+                if top_keyword not in competitor_topic_map:
+                    competitor_topic_map[top_keyword] = TopicGap(
+                        topic=top_keyword,
+                        estimated_traffic=int(page.get("traffic", 0)),
+                    )
+
+                gap = competitor_topic_map[top_keyword]
+                gap.competitor_urls.append(page.get("url", c_url))
+                gap.competitor_keywords.append(top_keyword)
+                gap.estimated_traffic = max(gap.estimated_traffic, int(page.get("traffic", 0)))
+
+        # Score gaps
+        gaps = list(competitor_topic_map.values())
+        for gap in gaps:
+            competitor_count = len(set(gap.competitor_urls))
+            traffic_score = min(100, math.log10(max(gap.estimated_traffic, 1)) / math.log10(10000) * 100)
+            competition_score = (competitor_count / max(len(competitor_urls), 1)) * 100
+            gap.priority_score = round((traffic_score * 0.6) + (competition_score * 0.4), 1)
+
+            # Difficulty estimation
+            if competitor_count >= 3:
+                gap.difficulty = "high"
+            elif competitor_count >= 2:
+                gap.difficulty = "medium"
+            else:
+                gap.difficulty = "low"
+
+            # Content type suggestion
+            gap.content_type_suggestion = self._suggest_content_type(gap.topic)
+
+        gaps.sort(key=lambda g: g.priority_score, reverse=True)
+        return gaps, target_keywords, content_volume
+
+    @staticmethod
+    def _suggest_content_type(topic: str) -> str:
+        """Suggest content type based on topic keywords."""
+        topic_lower = topic.lower()
+        if any(w in topic_lower for w in ["how to", "guide", "tutorial", "방법", "가이드"]):
+            return "guide"
+        if any(w in topic_lower for w in ["best", "top", "review", "추천", "후기", "비교"]):
+            return "listicle"
+        if any(w in topic_lower for w in ["what is", "이란", "뜻", "의미"]):
+            return "informational"
+        if any(w in topic_lower for w in ["cost", "price", "비용", "가격"]):
+            return "landing"
+        return "blog"
+
+    # ------------------------------------------------------------------
+    # Topic cluster mapping
+    # ------------------------------------------------------------------
+
+    def build_topic_clusters(
+        self,
+        topics: list[str],
+        n_clusters: int | None = None,
+        min_cluster_size: int = 3,
+    ) -> list[TopicCluster]:
+        """
+        Group topics into pillar/cluster structure using TF-IDF + hierarchical clustering.
+
+        Args:
+            topics: List of topic strings.
+            n_clusters: Number of clusters (auto-detected if None).
+            min_cluster_size: Minimum topics per cluster.
+
+        Returns:
+            List of TopicCluster objects.
+        """
+        if len(topics) < min_cluster_size:
+            self.logger.warning("Too few topics for clustering")
+            return []
+
+        # Vectorize topics
+        vectorizer = TfidfVectorizer(
+            max_features=500,
+            stop_words="english",
+            ngram_range=(1, 2),
+        )
+
+        try:
+            tfidf_matrix = vectorizer.fit_transform(topics)
+        except ValueError as exc:
+            self.logger.warning(f"TF-IDF vectorization failed: {exc}")
+            return []
+
+        # Auto-detect cluster count
+        if n_clusters is None:
+            n_clusters = max(2, min(len(topics) // 5, 15))
+
+        n_clusters = min(n_clusters, len(topics) - 1)
+
+        # Hierarchical clustering
+        clustering = AgglomerativeClustering(
+            n_clusters=n_clusters,
+            metric="cosine",
+            linkage="average",
+        )
+        labels = clustering.fit_predict(tfidf_matrix.toarray())
+
+        # Build cluster objects
+        cluster_map: dict[int, list[str]] = defaultdict(list)
+        for topic, label in zip(topics, labels):
+            cluster_map[label].append(topic)
+
+        clusters: list[TopicCluster] = []
+        for label, cluster_topics in sorted(cluster_map.items()):
+            if len(cluster_topics) < min_cluster_size:
+                continue
+
+            # Pick the longest topic as pillar (usually broader)
+            pillar = max(cluster_topics, key=len)
+            subtopics = [t for t in cluster_topics if t != pillar]
+
+            cluster = TopicCluster(
+                pillar_topic=pillar,
+                pillar_keyword=pillar,
+                cluster_topics=subtopics[:20],
+                cluster_keywords=[t for t in subtopics[:20]],
+                total_volume=0,
+                coverage_score=0.0,
+            )
+            clusters.append(cluster)
+
+        clusters.sort(key=lambda c: len(c.cluster_topics), reverse=True)
+        return clusters
+
+    # ------------------------------------------------------------------
+    # Editorial calendar generation
+    # ------------------------------------------------------------------
+
+    def generate_calendar(
+        self,
+        gaps: list[TopicGap],
+        clusters: list[TopicCluster],
+        weeks_ahead: int = 12,
+        entries_per_week: int = 2,
+    ) -> list[CalendarEntry]:
+        """
+        Generate prioritized editorial calendar from gaps and clusters.
+
+        Args:
+            gaps: List of topic gaps (sorted by priority).
+            clusters: List of topic clusters.
+            weeks_ahead: Number of weeks to plan.
+            entries_per_week: Content pieces per week.
+
+        Returns:
+            List of CalendarEntry objects.
+        """
+        calendar: list[CalendarEntry] = []
+        today = datetime.now()
+
+        # Build cluster lookup
+        topic_to_cluster: dict[str, str] = {}
+        for cluster in clusters:
+            for topic in cluster.cluster_topics:
+                topic_to_cluster[topic] = cluster.pillar_topic
+            topic_to_cluster[cluster.pillar_topic] = cluster.pillar_topic
+
+        # Prioritize: pillar topics first, then by priority score
+        pillar_topics = {c.pillar_topic for c in clusters}
+        pillar_gaps = [g for g in gaps if g.topic in pillar_topics]
+        other_gaps = [g for g in gaps if g.topic not in pillar_topics]
+        ordered_gaps = pillar_gaps + other_gaps
+
+        max_entries = weeks_ahead * entries_per_week
+        week_offset = 0
+        slot_in_week = 0
+
+        for gap in ordered_gaps[:max_entries]:
+            target_date = today + timedelta(weeks=week_offset, days=slot_in_week * 3)
+
+            # Determine priority label
+            if gap.priority_score >= 70:
+                priority = "high"
+            elif gap.priority_score >= 40:
+                priority = "medium"
+            else:
+                priority = "low"
+
+            # Word count based on content type
+            word_count_map = {
+                "guide": 2500,
+                "listicle": 2000,
+                "informational": 1800,
+                "landing": 1200,
+                "blog": 1500,
+            }
+
+            entry = CalendarEntry(
+                topic=gap.topic,
+                priority=priority,
+                target_date=target_date.strftime("%Y-%m-%d"),
+                content_type=gap.content_type_suggestion,
+                target_word_count=word_count_map.get(gap.content_type_suggestion, 1500),
+                primary_keyword=gap.topic,
+                estimated_traffic=gap.estimated_traffic,
+                cluster_name=topic_to_cluster.get(gap.topic, "uncategorized"),
+            )
+            calendar.append(entry)
+
+            slot_in_week += 1
+            if slot_in_week >= entries_per_week:
+                slot_in_week = 0
+                week_offset += 1
+
+        return calendar
+
+    # ------------------------------------------------------------------
+    # Korean opportunity detection
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def detect_korean_opportunities(gaps: list[TopicGap]) -> list[dict[str, Any]]:
+        """Detect Korean-market content opportunities in gaps."""
+        opportunities: list[dict[str, Any]] = []
+
+        for gap in gaps:
+            for pattern_info in KOREAN_OPPORTUNITY_PATTERNS:
+                if re.search(pattern_info["pattern"], gap.topic, re.IGNORECASE):
+                    opportunities.append({
+                        "topic": gap.topic,
+                        "pattern": pattern_info["label"],
+                        "description": pattern_info["description"],
+                        "estimated_traffic": gap.estimated_traffic,
+                        "priority_score": gap.priority_score,
+                    })
+                    break
+
+        opportunities.sort(key=lambda o: o["priority_score"], reverse=True)
+        return opportunities
+
+    # ------------------------------------------------------------------
+    # Orchestration
+    # ------------------------------------------------------------------
+
+    async def analyze(
+        self,
+        target_url: str,
+        competitor_urls: list[str],
+        build_clusters: bool = False,
+    ) -> ContentGapResult:
+        """
+        Run full content gap analysis.
+
+        Args:
+            target_url: Target website URL.
+            competitor_urls: List of competitor URLs.
+            build_clusters: Whether to build topic clusters.
+
+        Returns:
+            ContentGapResult with gaps, clusters, and calendar.
+        """
+        result = ContentGapResult(
+            target_url=target_url,
+            competitor_urls=competitor_urls,
+            timestamp=datetime.now().isoformat(),
+        )
+
+        self.logger.info(
+            f"Starting gap analysis: {target_url} vs {len(competitor_urls)} competitors"
+        )
+
+        # 1. Find topic gaps
+        gaps, target_keywords, content_volume = await self.find_topic_gaps(
+            target_url, competitor_urls
+        )
+
+        result.gaps = gaps
+        result.target_topics_count = len(target_keywords)
+        result.competitor_topics_count = sum(content_volume.get(c, 0) for c in competitor_urls)
+        result.content_volume_comparison = content_volume
+
+        # 2. Build topic clusters if requested
+        if build_clusters and gaps:
+            all_topics = [g.topic for g in gaps]
+            result.clusters = self.build_topic_clusters(all_topics)
+
+        # 3. Generate editorial calendar
+        result.calendar = self.generate_calendar(gaps, result.clusters)
+
+        # 4. Detect Korean opportunities
+        result.korean_opportunities = self.detect_korean_opportunities(gaps)
+
+        # 5. Recommendations
+        result.recommendations = self._generate_recommendations(result)
+
+        self.logger.info(
+            f"Gap analysis complete: {len(gaps)} gaps, "
+            f"{len(result.clusters)} clusters, "
+            f"{len(result.calendar)} calendar entries"
+        )
+
+        return result
+
+    @staticmethod
+    def _generate_recommendations(result: ContentGapResult) -> list[str]:
+        """Generate strategic recommendations from gap analysis."""
+        recs: list[str] = []
+
+        gap_count = len(result.gaps)
+        if gap_count > 50:
+            recs.append(
+                f"경쟁사 대비 {gap_count}개의 콘텐츠 격차가 발견되었습니다. "
+                "우선순위 상위 20개 주제부터 콘텐츠 생성을 시작하세요."
+            )
+        elif gap_count > 20:
+            recs.append(
+                f"{gap_count}개의 콘텐츠 격차가 있습니다. "
+                "높은 트래픽 기회부터 순차적으로 콘텐츠를 생성하세요."
+            )
+        elif gap_count > 0:
+            recs.append(
+                f"{gap_count}개의 콘텐츠 격차가 발견되었습니다. "
+                "비교적 적은 격차이므로 빠른 시일 내 모두 커버할 수 있습니다."
+            )
+
+        if result.clusters:
+            recs.append(
+                f"{len(result.clusters)}개의 토픽 클러스터를 구성했습니다. "
+                "필러 콘텐츠부터 작성하여 내부 링크 구조를 강화하세요."
+            )
+
+        if result.korean_opportunities:
+            recs.append(
+                f"한국어 시장 기회가 {len(result.korean_opportunities)}개 발견되었습니다. "
+                "후기, 비용, 비교 콘텐츠는 한국 검색 시장에서 높은 전환율을 보입니다."
+            )
+
+        high_priority = [g for g in result.gaps if g.priority_score >= 70]
+        if high_priority:
+            top_topics = ", ".join(g.topic for g in high_priority[:3])
+            recs.append(
+                f"최우선 주제: {top_topics}. "
+                "이 주제들은 높은 트래픽 잠재력과 경쟁사 커버리지를 가지고 있습니다."
+            )
+
+        if not recs:
+            recs.append("경쟁사 대비 콘텐츠 커버리지가 양호합니다. 기존 콘텐츠 최적화에 집중하세요.")
+
+        return recs
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="SEO Content Gap Analyzer - topic gaps, clusters, calendar",
+    )
+    parser.add_argument("--target", required=True, help="Target website URL")
+    parser.add_argument(
+        "--competitor", action="append", dest="competitors", required=True,
+        help="Competitor URL (can be repeated)",
+    )
+    parser.add_argument("--clusters", action="store_true", help="Build topic clusters")
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", help="Save output to file")
+    return parser
+
+
+def format_text_report(result: ContentGapResult) -> str:
+    """Format gap analysis result as human-readable text."""
+    lines: list[str] = []
+    lines.append(f"## Content Gap Analysis: {result.target_url}")
+    lines.append(f"**Date**: {result.timestamp[:10]}")
+    lines.append(f"**Competitors**: {', '.join(result.competitor_urls)}")
+    lines.append("")
+
+    lines.append("### Content Volume Comparison")
+    for site, count in result.content_volume_comparison.items():
+        lines.append(f"  - {site}: {count} topics")
+    lines.append("")
+
+    lines.append(f"### Topic Gaps ({len(result.gaps)} found)")
+    for i, gap in enumerate(result.gaps[:20], 1):
+        lines.append(
+            f"  {i}. [{gap.priority_score:.0f}] {gap.topic} "
+            f"(traffic: {gap.estimated_traffic}, difficulty: {gap.difficulty})"
+        )
+    lines.append("")
+
+    if result.clusters:
+        lines.append(f"### Topic Clusters ({len(result.clusters)})")
+        for i, cluster in enumerate(result.clusters, 1):
+            lines.append(f"  {i}. **{cluster.pillar_topic}** ({len(cluster.cluster_topics)} subtopics)")
+            for sub in cluster.cluster_topics[:5]:
+                lines.append(f"     - {sub}")
+        lines.append("")
+
+    if result.calendar:
+        lines.append(f"### Editorial Calendar ({len(result.calendar)} entries)")
+        for entry in result.calendar[:15]:
+            lines.append(
+                f"  - [{entry.target_date}] {entry.topic} "
+                f"({entry.content_type}, {entry.target_word_count}w, priority: {entry.priority})"
+            )
+        lines.append("")
+
+    if result.korean_opportunities:
+        lines.append(f"### Korean Market Opportunities ({len(result.korean_opportunities)})")
+        for opp in result.korean_opportunities[:10]:
+            lines.append(f"  - {opp['topic']} ({opp['description']})")
+        lines.append("")
+
+    lines.append("### Recommendations")
+    for i, rec in enumerate(result.recommendations, 1):
+        lines.append(f"  {i}. {rec}")
+
+    return "\n".join(lines)
+
+
+async def main() -> None:
+    parser = build_parser()
+    args = parser.parse_args()
+
+    analyzer = ContentGapAnalyzer()
+    result = await analyzer.analyze(
+        target_url=args.target,
+        competitor_urls=args.competitors,
+        build_clusters=args.clusters,
+    )
+
+    if args.json:
+        output = json.dumps(asdict(result), ensure_ascii=False, indent=2, default=str)
+    else:
+        output = format_text_report(result)
+
+    if args.output:
+        with open(args.output, "w", encoding="utf-8") as f:
+            f.write(output)
+        logger.info(f"Output saved to {args.output}")
+    else:
+        print(output)
+
+    analyzer.print_stats()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/23-seo-content-strategy/code/scripts/requirements.txt b/custom-skills/23-seo-content-strategy/code/scripts/requirements.txt
new file mode 100644
index 0000000..173019f
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/code/scripts/requirements.txt
@@ -0,0 +1,11 @@
+# 23-seo-content-strategy dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+beautifulsoup4>=4.12.0
+lxml>=5.1.0
+pandas>=2.1.0
+scikit-learn>=1.3.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/23-seo-content-strategy/desktop/SKILL.md b/custom-skills/23-seo-content-strategy/desktop/SKILL.md
new file mode 100644
index 0000000..53f093c
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/desktop/SKILL.md
@@ -0,0 +1,138 @@
+---
+name: seo-content-strategy
+description: |
+  Content strategy and planning for SEO. Triggers: content audit, content strategy, content gap, topic clusters, content brief, editorial calendar, content decay, 콘텐츠 전략, 콘텐츠 감사.
+---
+
+# SEO Content Strategy
+
+## Purpose
+
+Audit existing content performance, identify topic gaps vs competitors, map topic clusters, detect content decay, and generate SEO content briefs. Supports Korean content patterns (Naver Blog format, 후기/review content, 추천 listicles).
+
+## Core Capabilities
+
+1. **Content Audit** - Inventory, performance scoring, decay detection
+2. **Content Gap Analysis** - Topic gaps vs competitors, cluster mapping
+3. **Content Brief Generation** - Outlines, keywords, word count targets
+4. **Editorial Calendar** - Prioritized content creation schedule
+5. **Korean Content Patterns** - Naver Blog style, 후기, 추천 format analysis
+
+## MCP Tool Usage
+
+### Ahrefs for Content Data
+```
+site-explorer-top-pages: Get top performing pages
+site-explorer-pages-by-traffic: Pages ranked by organic traffic
+site-explorer-organic-keywords: Keywords per page
+site-explorer-organic-competitors: Find content competitors
+site-explorer-best-by-external-links: Best content by backlinks
+keywords-explorer-matching-terms: Secondary keyword suggestions
+keywords-explorer-related-terms: LSI keyword suggestions
+serp-overview: Analyze top ranking results for a keyword
+```
+
+### WebSearch for Content Research
+```
+WebSearch: Research content topics and competitor strategies
+WebFetch: Analyze competitor page content and structure
+```
+
+### Notion for Report Storage
+```
+notion-create-pages: Save audit reports to SEO Audit Log
+```
+
+## Workflow
+
+### 1. Content Audit
+1. Crawl sitemap to discover all content URLs
+2. Fetch top pages data from Ahrefs (traffic, keywords, backlinks)
+3. Classify content types (blog, product, service, landing, resource)
+4. Score each page performance (0-100 composite)
+5. Detect decaying content (traffic decline patterns)
+6. Analyze freshness distribution (fresh/aging/stale)
+7. Identify Korean content patterns (후기, 추천, 방법 formats)
+8. Generate recommendations
+
+### 2. Content Gap Analysis
+1. Gather target site keywords from Ahrefs
+2. Gather competitor top pages and keywords
+3. Identify topics present in competitors but missing from target
+4. Score gaps by priority (traffic potential + competition coverage)
+5. Build topic clusters using TF-IDF + hierarchical clustering
+6. Generate editorial calendar with priority and dates
+7. Detect Korean market content opportunities
+
+### 3. Content Brief Generation
+1. Analyze top 5-10 ranking pages for target keyword
+2. Extract headings, word counts, content features (FAQ, images, video)
+3. Build recommended H2/H3 outline from competitor patterns
+4. Suggest primary, secondary, and LSI keywords
+5. Calculate target word count (avg of top 5 +/- 20%)
+6. Find internal linking opportunities on the target site
+7. Detect search intent (informational, commercial, transactional, navigational)
+8. Add Korean format recommendations based on intent
+
+## Output Format
+
+```markdown
+## Content Audit: [domain]
+
+### Content Inventory
+- Total pages: [count]
+- By type: blog [n], product [n], service [n], other [n]
+- Average performance score: [score]/100
+
+### Top Performers
+1. [score] [url] (traffic: [n])
+...
+
+### Decaying Content
+1. [decay rate] [url] (traffic: [n])
+...
+
+### Content Gaps vs Competitors
+1. [priority] [topic] (est. traffic: [n], difficulty: [level])
+...
+
+### Topic Clusters
+1. **[Pillar Topic]** ([n] subtopics)
+   - [subtopic 1]
+   - [subtopic 2]
+
+### Editorial Calendar
+- [date] [topic] ([type], [word count], priority: [level])
+...
+
+### Recommendations
+1. [Priority actions]
+```
+
+## Common Issues
+
+| Issue | Impact | Fix |
+|-------|--------|-----|
+| No blog content | High | Build blog content strategy with topic clusters |
+| Content decay (traffic loss) | High | Refresh and update declining pages |
+| Missing competitor topics | Medium | Create content for high-priority gaps |
+| No 후기/review content | Medium | Add Korean review-style content for conversions |
+| Stale content (>12 months) | Medium | Update or consolidate outdated pages |
+| No topic clusters | Medium | Organize content into pillar/cluster structure |
+| Missing FAQ sections | Low | Add FAQ schema for featured snippet opportunities |
+
+## Limitations
+
+- Ahrefs API required for traffic and keyword data
+- Competitor analysis limited to publicly available content
+- Content decay detection uses heuristic without historical data in standalone mode
+- Topic clustering requires minimum 3 topics per cluster
+- Word count analysis requires accessible competitor pages (no JS rendering)
+
+## Notion Output (Required)
+
+All audit reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category, Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: CONTENT-YYYYMMDD-NNN
diff --git a/custom-skills/23-seo-content-strategy/desktop/skill.yaml b/custom-skills/23-seo-content-strategy/desktop/skill.yaml
new file mode 100644
index 0000000..4ec2631
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-content-strategy
+description: |
+  Content strategy and planning for SEO. Triggers: content audit, content strategy, content gap, topic clusters, content brief, editorial calendar, content decay.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/23-seo-content-strategy/desktop/tools/ahrefs.md b/custom-skills/23-seo-content-strategy/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..5d918e0
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/desktop/tools/ahrefs.md
@@ -0,0 +1,15 @@
+# Ahrefs
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/23-seo-content-strategy/desktop/tools/notion.md b/custom-skills/23-seo-content-strategy/desktop/tools/notion.md
new file mode 100644
index 0000000..153e086
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/desktop/tools/notion.md
@@ -0,0 +1,15 @@
+# Notion
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/23-seo-content-strategy/desktop/tools/websearch.md b/custom-skills/23-seo-content-strategy/desktop/tools/websearch.md
new file mode 100644
index 0000000..36ef238
--- /dev/null
+++ b/custom-skills/23-seo-content-strategy/desktop/tools/websearch.md
@@ -0,0 +1,15 @@
+# WebSearch
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/24-seo-ecommerce/code/CLAUDE.md b/custom-skills/24-seo-ecommerce/code/CLAUDE.md
new file mode 100644
index 0000000..0c2be5a
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/code/CLAUDE.md
@@ -0,0 +1,136 @@
+# CLAUDE.md
+
+## Overview
+
+E-commerce SEO audit tool for product page optimization, product schema validation, category taxonomy analysis, and marketplace presence checking. Supports Naver Smart Store optimization and Korean marketplace platforms (Coupang, Gmarket, 11번가).
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# E-commerce SEO audit
+python scripts/ecommerce_auditor.py --url https://example.com --json
+
+# Product schema validation
+python scripts/product_schema_checker.py --url https://example.com/product --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `ecommerce_auditor.py` | Full e-commerce SEO audit | Product page issues, category structure, marketplace presence |
+| `product_schema_checker.py` | Validate product structured data | Schema completeness, errors, rich result eligibility |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## E-Commerce Auditor
+
+```bash
+# Full audit
+python scripts/ecommerce_auditor.py --url https://example.com --json
+
+# Product page audit only
+python scripts/ecommerce_auditor.py --url https://example.com --scope products --json
+
+# Category taxonomy analysis
+python scripts/ecommerce_auditor.py --url https://example.com --scope categories --json
+
+# Check Korean marketplace presence
+python scripts/ecommerce_auditor.py --url https://example.com --korean-marketplaces --json
+```
+
+**Capabilities**:
+- Product page SEO audit (titles, meta descriptions, image alt text, H1 structure)
+- Category taxonomy analysis (depth, breadcrumb implementation, faceted navigation)
+- Duplicate content detection (parameter URLs, product variants, pagination)
+- Pagination/infinite scroll SEO validation (rel=prev/next, canonical tags)
+- Internal linking structure for product discovery
+- Naver Smart Store optimization checks
+- Korean marketplace presence (Coupang, Gmarket, 11번가 product listing detection)
+
+## Product Schema Checker
+
+```bash
+# Validate single product page
+python scripts/product_schema_checker.py --url https://example.com/product/123 --json
+
+# Batch validate from sitemap
+python scripts/product_schema_checker.py --sitemap https://example.com/product-sitemap.xml --sample 50 --json
+```
+
+**Capabilities**:
+- Product schema validation (Product, Offer, AggregateRating, Review, BreadcrumbList)
+- Required property completeness check (name, image, description, offers, price, availability)
+- Optional property recommendations (brand, sku, gtin, mpn, review, aggregateRating)
+- Rich result eligibility assessment
+- Price and availability markup validation
+- Merchant listing schema support
+- Korean market: Naver Shopping structured data requirements
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `site-explorer-pages-by-traffic` | Identify top product/category pages |
+| `site-explorer-organic-keywords` | Product page keyword performance |
+
+## Output Format
+
+```json
+{
+  "url": "https://example.com",
+  "product_pages_audited": 50,
+  "issues": {
+    "critical": [...],
+    "high": [...],
+    "medium": [...],
+    "low": [...]
+  },
+  "category_structure": {
+    "max_depth": 4,
+    "breadcrumbs_present": true,
+    "faceted_nav_issues": [...]
+  },
+  "schema_validation": {
+    "pages_with_schema": 42,
+    "pages_without_schema": 8,
+    "common_errors": [...]
+  },
+  "korean_marketplaces": {
+    "naver_smart_store": {"found": true, "url": "..."},
+    "coupang": {"found": false},
+    "gmarket": {"found": false}
+  },
+  "score": 65,
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Audited website URL |
+| Category | Select | E-Commerce SEO |
+| Priority | Select | Based on issue severity |
+| Found Date | Date | Audit date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: ECOM-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., Schema Markup, Product, Offer)
+- URLs and code remain unchanged
diff --git a/custom-skills/24-seo-ecommerce/code/scripts/base_client.py b/custom-skills/24-seo-ecommerce/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/24-seo-ecommerce/code/scripts/ecommerce_auditor.py b/custom-skills/24-seo-ecommerce/code/scripts/ecommerce_auditor.py
new file mode 100644
index 0000000..b38e4a2
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/code/scripts/ecommerce_auditor.py
@@ -0,0 +1,1046 @@
+"""
+E-Commerce SEO Auditor
+======================
+Purpose: Audit product pages, category taxonomy, duplicate content,
+         pagination SEO, and Korean marketplace presence.
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import re
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from typing import Any
+from urllib.parse import parse_qs, urljoin, urlparse
+
+import aiohttp
+from bs4 import BeautifulSoup
+from rich.console import Console
+from rich.table import Table
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+console = Console()
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class ProductPageIssue:
+    """Single issue found on a product page."""
+    url: str
+    issue_type: str          # title, meta_desc, h1, image_alt, internal_link, canonical, pagination
+    severity: str            # critical, high, medium, low
+    message: str
+    recommendation: str
+
+
+@dataclass
+class CategoryNode:
+    """A node in the category taxonomy tree."""
+    url: str
+    name: str
+    depth: int
+    children_count: int
+    has_breadcrumb: bool
+
+
+@dataclass
+class DuplicateGroup:
+    """Group of duplicate or near-duplicate product URLs."""
+    canonical_url: str
+    duplicate_urls: list[str]
+    reason: str              # parameter_variant, product_variant, pagination_missing_canonical
+
+
+@dataclass
+class MarketplacePresence:
+    """Presence record for a Korean marketplace."""
+    platform: str            # naver_smart_store, coupang, gmarket, 11st
+    found: bool
+    url: str | None = None
+    product_count: int = 0
+
+
+@dataclass
+class EcommerceAuditResult:
+    """Complete e-commerce SEO audit result."""
+    url: str
+    product_pages_audited: int = 0
+    issues: dict[str, list[dict]] = field(default_factory=lambda: {
+        "critical": [], "high": [], "medium": [], "low": []
+    })
+    category_structure: dict[str, Any] = field(default_factory=dict)
+    duplicate_groups: list[dict] = field(default_factory=list)
+    pagination_issues: list[dict] = field(default_factory=list)
+    korean_marketplaces: dict[str, dict] = field(default_factory=dict)
+    naver_smart_store: dict[str, Any] = field(default_factory=dict)
+    score: int = 0
+    timestamp: str = ""
+
+    def add_issue(self, issue: ProductPageIssue) -> None:
+        self.issues[issue.severity].append(asdict(issue))
+
+    def calculate_score(self) -> int:
+        """Score 0-100 based on issue severity counts."""
+        penalties = {
+            "critical": 15,
+            "high": 8,
+            "medium": 3,
+            "low": 1,
+        }
+        total_penalty = sum(
+            len(items) * penalties[sev]
+            for sev, items in self.issues.items()
+        )
+        self.score = max(0, 100 - total_penalty)
+        return self.score
+
+
+# ---------------------------------------------------------------------------
+# Product URL pattern helpers
+# ---------------------------------------------------------------------------
+
+PRODUCT_URL_PATTERNS = [
+    r"/product[s]?/",
+    r"/item[s]?/",
+    r"/p/",
+    r"/dp/",
+    r"/goods/",
+    r"/shop/",
+    r"/detail/",
+    r"\?product_id=",
+    r"\?item_id=",
+    r"\?goodsno=",
+]
+
+CATEGORY_URL_PATTERNS = [
+    r"/category/",
+    r"/categories/",
+    r"/collections?/",
+    r"/c/",
+    r"/department/",
+    r"/browse/",
+]
+
+FACETED_NAV_PARAMS = [
+    "color", "size", "sort", "order", "filter", "brand",
+    "price_min", "price_max", "page", "per_page", "view",
+    "material", "rating", "availability",
+]
+
+
+def is_product_url(url: str) -> bool:
+    """Check if a URL looks like a product page."""
+    return any(re.search(pat, url, re.IGNORECASE) for pat in PRODUCT_URL_PATTERNS)
+
+
+def is_category_url(url: str) -> bool:
+    """Check if a URL looks like a category page."""
+    return any(re.search(pat, url, re.IGNORECASE) for pat in CATEGORY_URL_PATTERNS)
+
+
+def get_faceted_params(url: str) -> dict[str, list[str]]:
+    """Extract faceted navigation parameters from a URL."""
+    parsed = urlparse(url)
+    params = parse_qs(parsed.query)
+    return {k: v for k, v in params.items() if k.lower() in FACETED_NAV_PARAMS}
+
+
+# ---------------------------------------------------------------------------
+# Main auditor
+# ---------------------------------------------------------------------------
+
+class EcommerceAuditor(BaseAsyncClient):
+    """E-commerce SEO auditor with product, category, and marketplace checks."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 10,
+        requests_per_second: float = 5.0,
+        timeout: int = 30,
+    ):
+        super().__init__(max_concurrent=max_concurrent, requests_per_second=requests_per_second)
+        self.timeout = aiohttp.ClientTimeout(total=timeout)
+        self.headers = {
+            "User-Agent": (
+                "Mozilla/5.0 (compatible; EcommerceSEOBot/1.0; "
+                "+https://ourdigital.org)"
+            ),
+            "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
+            "Accept-Language": "ko-KR,ko;q=0.9,en-US;q=0.8,en;q=0.7",
+        }
+
+    # ------------------------------------------------------------------
+    # Page fetching
+    # ------------------------------------------------------------------
+
+    async def _fetch_page(self, session: aiohttp.ClientSession, url: str) -> tuple[str, str]:
+        """Fetch a page and return (final_url, html)."""
+        try:
+            async with session.get(url, headers=self.headers, timeout=self.timeout,
+                                   allow_redirects=True, ssl=False) as resp:
+                html = await resp.text(errors="replace")
+                return str(resp.url), html
+        except Exception as exc:
+            self.logger.warning(f"Failed to fetch {url}: {exc}")
+            return url, ""
+
+    # ------------------------------------------------------------------
+    # Product page discovery via Ahrefs
+    # ------------------------------------------------------------------
+
+    async def get_product_pages(self, domain: str, sample: int = 50) -> list[dict]:
+        """
+        Discover product pages using Ahrefs pages-by-traffic.
+        Falls back to sitemap crawling if Ahrefs is unavailable.
+        Returns list of dicts with keys: url, traffic, keywords.
+        """
+        pages: list[dict] = []
+
+        # Attempt Ahrefs via environment / MCP
+        self.logger.info(f"Discovering product pages for {domain} (sample={sample})")
+
+        # Fallback: fetch sitemap and filter product URLs
+        sitemap_urls = await self._fetch_sitemap_urls(domain)
+        product_urls = [u for u in sitemap_urls if is_product_url(u)]
+        if not product_urls:
+            # Broader heuristic: any URL with 3+ path segments
+            product_urls = [
+                u for u in sitemap_urls
+                if len(urlparse(u).path.strip("/").split("/")) >= 2
+            ]
+
+        for url in product_urls[:sample]:
+            pages.append({"url": url, "traffic": 0, "keywords": 0})
+
+        self.logger.info(f"Found {len(pages)} product page candidates")
+        return pages
+
+    async def _fetch_sitemap_urls(self, domain: str) -> list[str]:
+        """Fetch URLs from the site's XML sitemap."""
+        urls: list[str] = []
+        base = f"https://{domain}" if not domain.startswith("http") else domain
+        parsed = urlparse(base)
+        sitemap_url = f"{parsed.scheme}://{parsed.netloc}/sitemap.xml"
+
+        async with aiohttp.ClientSession() as session:
+            try:
+                async with session.get(sitemap_url, headers=self.headers,
+                                       timeout=self.timeout, ssl=False) as resp:
+                    if resp.status == 200:
+                        text = await resp.text(errors="replace")
+                        soup = BeautifulSoup(text, "lxml-xml")
+                        # Handle sitemap index
+                        sitemapindex = soup.find_all("sitemap")
+                        if sitemapindex:
+                            for sm in sitemapindex[:5]:
+                                loc = sm.find("loc")
+                                if loc:
+                                    child_urls = await self._parse_sitemap(session, loc.text.strip())
+                                    urls.extend(child_urls)
+                        else:
+                            url_tags = soup.find_all("url")
+                            for tag in url_tags:
+                                loc = tag.find("loc")
+                                if loc:
+                                    urls.append(loc.text.strip())
+            except Exception as exc:
+                self.logger.warning(f"Sitemap fetch failed: {exc}")
+
+        return urls
+
+    async def _parse_sitemap(self, session: aiohttp.ClientSession, url: str) -> list[str]:
+        """Parse a single sitemap XML file and return its URLs."""
+        urls: list[str] = []
+        try:
+            async with session.get(url, headers=self.headers,
+                                   timeout=self.timeout, ssl=False) as resp:
+                if resp.status == 200:
+                    text = await resp.text(errors="replace")
+                    soup = BeautifulSoup(text, "lxml-xml")
+                    for tag in soup.find_all("url"):
+                        loc = tag.find("loc")
+                        if loc:
+                            urls.append(loc.text.strip())
+        except Exception as exc:
+            self.logger.warning(f"Failed to parse sitemap {url}: {exc}")
+        return urls
+
+    # ------------------------------------------------------------------
+    # Product page audit
+    # ------------------------------------------------------------------
+
+    async def audit_product_page(
+        self,
+        session: aiohttp.ClientSession,
+        page_url: str,
+    ) -> list[ProductPageIssue]:
+        """Audit a single product page for SEO issues."""
+        issues: list[ProductPageIssue] = []
+        _, html = await self._fetch_page(session, page_url)
+        if not html:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="accessibility", severity="critical",
+                message="Page returned empty or could not be fetched",
+                recommendation="Verify the URL is accessible and returns valid HTML.",
+            ))
+            return issues
+
+        soup = BeautifulSoup(html, "lxml")
+
+        # --- Title tag ---
+        title_tag = soup.find("title")
+        title_text = title_tag.get_text(strip=True) if title_tag else ""
+        if not title_text:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="title", severity="critical",
+                message="Missing <title> tag",
+                recommendation="Add a unique title containing the product name (under 60 characters).",
+            ))
+        elif len(title_text) > 60:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="title", severity="medium",
+                message=f"Title too long ({len(title_text)} chars): {title_text[:80]}...",
+                recommendation="Shorten title to under 60 characters for full SERP display.",
+            ))
+        elif len(title_text) < 15:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="title", severity="medium",
+                message=f"Title too short ({len(title_text)} chars): {title_text}",
+                recommendation="Expand title with product name, key feature, and brand.",
+            ))
+
+        # --- Meta description ---
+        meta_desc_tag = soup.find("meta", attrs={"name": re.compile(r"description", re.I)})
+        meta_desc = meta_desc_tag.get("content", "").strip() if meta_desc_tag else ""
+        if not meta_desc:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="meta_desc", severity="high",
+                message="Missing meta description",
+                recommendation="Add meta description with product features and price info (under 155 chars).",
+            ))
+        elif len(meta_desc) > 155:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="meta_desc", severity="low",
+                message=f"Meta description too long ({len(meta_desc)} chars)",
+                recommendation="Trim to under 155 characters for full SERP display.",
+            ))
+
+        # --- H1 tag ---
+        h1_tags = soup.find_all("h1")
+        if not h1_tags:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="h1", severity="high",
+                message="Missing H1 tag on product page",
+                recommendation="Add a single H1 with the product name.",
+            ))
+        elif len(h1_tags) > 1:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="h1", severity="medium",
+                message=f"Multiple H1 tags found ({len(h1_tags)})",
+                recommendation="Use a single H1 for the product name; use H2/H3 for subsections.",
+            ))
+
+        # --- Image alt text ---
+        images = soup.find_all("img")
+        product_images = [
+            img for img in images
+            if img.get("src") and not any(
+                skip in (img.get("src", "") + img.get("class", [""])[0] if img.get("class") else img.get("src", ""))
+                for skip in ["logo", "icon", "badge", "banner", "sprite", "pixel", "tracking"]
+            )
+        ]
+        missing_alt = [img for img in product_images if not img.get("alt", "").strip()]
+        if missing_alt:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="image_alt", severity="high",
+                message=f"{len(missing_alt)} product image(s) missing alt text",
+                recommendation="Add descriptive alt text with product name to all product images.",
+            ))
+
+        generic_alt = [
+            img for img in product_images
+            if img.get("alt", "").strip().lower() in [
+                "image", "photo", "product", "picture", "img", "product image",
+                "상품 이미지", "이미지", "사진",
+            ]
+        ]
+        if generic_alt:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="image_alt", severity="medium",
+                message=f"{len(generic_alt)} image(s) with generic alt text",
+                recommendation="Replace generic alt text with specific product descriptions.",
+            ))
+
+        # --- Canonical tag ---
+        canonical = soup.find("link", attrs={"rel": "canonical"})
+        if not canonical:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="canonical", severity="high",
+                message="Missing canonical tag on product page",
+                recommendation="Add <link rel='canonical'> pointing to the preferred product URL.",
+            ))
+        else:
+            canonical_href = canonical.get("href", "").strip()
+            if canonical_href and canonical_href != page_url:
+                # Only flag if significantly different (not just trailing slash)
+                norm_canonical = canonical_href.rstrip("/")
+                norm_page = page_url.rstrip("/")
+                if norm_canonical != norm_page:
+                    issues.append(ProductPageIssue(
+                        url=page_url, issue_type="canonical", severity="medium",
+                        message=f"Canonical points to different URL: {canonical_href}",
+                        recommendation="Verify canonical is correct; ensure product variants point to the main product.",
+                    ))
+
+        # --- Internal links ---
+        internal_links = []
+        parsed_page = urlparse(page_url)
+        for a_tag in soup.find_all("a", href=True):
+            href = a_tag["href"]
+            full_url = urljoin(page_url, href)
+            parsed_link = urlparse(full_url)
+            if parsed_link.netloc == parsed_page.netloc:
+                internal_links.append(full_url)
+
+        if len(internal_links) < 3:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="internal_link", severity="medium",
+                message=f"Only {len(internal_links)} internal links found",
+                recommendation="Add related product links, category breadcrumbs, and cross-sell sections.",
+            ))
+
+        # --- Open Graph / social meta ---
+        og_title = soup.find("meta", attrs={"property": "og:title"})
+        og_image = soup.find("meta", attrs={"property": "og:image"})
+        if not og_title or not og_image:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="social_meta", severity="low",
+                message="Missing Open Graph tags (og:title or og:image)",
+                recommendation="Add OG tags for better social sharing of product pages.",
+            ))
+
+        return issues
+
+    # ------------------------------------------------------------------
+    # Category taxonomy analysis
+    # ------------------------------------------------------------------
+
+    async def analyze_category_taxonomy(
+        self,
+        session: aiohttp.ClientSession,
+        base_url: str,
+        max_categories: int = 50,
+    ) -> dict[str, Any]:
+        """Analyze category page structure and taxonomy depth."""
+        result: dict[str, Any] = {
+            "categories_found": 0,
+            "max_depth": 0,
+            "avg_depth": 0.0,
+            "breadcrumbs_present": 0,
+            "breadcrumbs_missing": 0,
+            "faceted_nav_issues": [],
+            "nodes": [],
+        }
+
+        # Discover category URLs from sitemap
+        sitemap_urls = await self._fetch_sitemap_urls(base_url)
+        category_urls = [u for u in sitemap_urls if is_category_url(u)][:max_categories]
+
+        if not category_urls:
+            # Try crawling homepage for category links
+            _, html = await self._fetch_page(session, base_url)
+            if html:
+                soup = BeautifulSoup(html, "lxml")
+                for a_tag in soup.find_all("a", href=True):
+                    full_url = urljoin(base_url, a_tag["href"])
+                    if is_category_url(full_url) and full_url not in category_urls:
+                        category_urls.append(full_url)
+                        if len(category_urls) >= max_categories:
+                            break
+
+        depths: list[int] = []
+        for cat_url in category_urls:
+            _, html = await self._fetch_page(session, cat_url)
+            if not html:
+                continue
+
+            soup = BeautifulSoup(html, "lxml")
+            parsed = urlparse(cat_url)
+            path_parts = [p for p in parsed.path.strip("/").split("/") if p]
+            depth = len(path_parts)
+            depths.append(depth)
+
+            # Check breadcrumb
+            has_breadcrumb = bool(
+                soup.find("nav", attrs={"aria-label": re.compile(r"breadcrumb", re.I)})
+                or soup.find(attrs={"class": re.compile(r"breadcrumb", re.I)})
+                or soup.find("script", string=re.compile(r"BreadcrumbList", re.I))
+                or soup.find("ol", attrs={"itemtype": re.compile(r"BreadcrumbList", re.I)})
+            )
+
+            if has_breadcrumb:
+                result["breadcrumbs_present"] += 1
+            else:
+                result["breadcrumbs_missing"] += 1
+
+            # Category name from H1 or title
+            h1 = soup.find("h1")
+            cat_name = h1.get_text(strip=True) if h1 else path_parts[-1] if path_parts else "unknown"
+
+            # Count child category links
+            children = 0
+            for a_tag in soup.find_all("a", href=True):
+                link = urljoin(cat_url, a_tag["href"])
+                if is_category_url(link) and link != cat_url:
+                    children += 1
+
+            node = CategoryNode(
+                url=cat_url,
+                name=cat_name,
+                depth=depth,
+                children_count=children,
+                has_breadcrumb=has_breadcrumb,
+            )
+            result["nodes"].append(asdict(node))
+
+            # Faceted navigation check
+            faceted = get_faceted_params(cat_url)
+            if faceted:
+                robots_meta = soup.find("meta", attrs={"name": "robots"})
+                robots_content = robots_meta.get("content", "").lower() if robots_meta else ""
+                canonical = soup.find("link", attrs={"rel": "canonical"})
+                canonical_href = canonical.get("href", "").strip() if canonical else ""
+
+                if "noindex" not in robots_content and canonical_href == cat_url:
+                    result["faceted_nav_issues"].append({
+                        "url": cat_url,
+                        "params": list(faceted.keys()),
+                        "message": "Faceted URL is indexable without canonical to base category",
+                        "recommendation": (
+                            "Add noindex or canonical to the non-filtered category URL "
+                            "to prevent duplicate content."
+                        ),
+                    })
+
+        if depths:
+            result["max_depth"] = max(depths)
+            result["avg_depth"] = round(sum(depths) / len(depths), 1)
+        result["categories_found"] = len(category_urls)
+
+        return result
+
+    # ------------------------------------------------------------------
+    # Duplicate content detection
+    # ------------------------------------------------------------------
+
+    async def detect_duplicates(self, page_urls: list[str]) -> list[DuplicateGroup]:
+        """Detect potential duplicate content from parameter variants."""
+        groups: list[DuplicateGroup] = []
+        base_to_variants: dict[str, list[str]] = {}
+
+        for url in page_urls:
+            parsed = urlparse(url)
+            base = f"{parsed.scheme}://{parsed.netloc}{parsed.path}"
+            params = parse_qs(parsed.query)
+
+            faceted = {k: v for k, v in params.items() if k.lower() in FACETED_NAV_PARAMS}
+            if faceted:
+                base_to_variants.setdefault(base, []).append(url)
+
+        for base_url, variants in base_to_variants.items():
+            if len(variants) > 1:
+                groups.append(DuplicateGroup(
+                    canonical_url=base_url,
+                    duplicate_urls=variants,
+                    reason="parameter_variant",
+                ))
+
+        # Check for product variant duplicates (e.g., /product/123-red vs /product/123-blue)
+        slug_groups: dict[str, list[str]] = {}
+        for url in page_urls:
+            parsed = urlparse(url)
+            path = parsed.path.rstrip("/")
+            # Strip trailing color/size suffixes
+            base_slug = re.sub(
+                r"[-_](red|blue|green|black|white|small|medium|large|xs|s|m|l|xl|xxl)$",
+                "", path, flags=re.IGNORECASE,
+            )
+            if base_slug != path:
+                slug_groups.setdefault(base_slug, []).append(url)
+
+        for base_slug, variants in slug_groups.items():
+            if len(variants) > 1:
+                groups.append(DuplicateGroup(
+                    canonical_url=variants[0],
+                    duplicate_urls=variants[1:],
+                    reason="product_variant",
+                ))
+
+        return groups
+
+    # ------------------------------------------------------------------
+    # Pagination SEO
+    # ------------------------------------------------------------------
+
+    async def check_pagination_seo(
+        self,
+        session: aiohttp.ClientSession,
+        page_url: str,
+    ) -> list[ProductPageIssue]:
+        """Check pagination implementation for SEO best practices."""
+        issues: list[ProductPageIssue] = []
+        _, html = await self._fetch_page(session, page_url)
+        if not html:
+            return issues
+
+        soup = BeautifulSoup(html, "lxml")
+
+        # Look for pagination links
+        pagination_links = []
+        for a_tag in soup.find_all("a", href=True):
+            href = a_tag["href"]
+            full_url = urljoin(page_url, href)
+            params = parse_qs(urlparse(full_url).query)
+            if "page" in params or "p" in params or re.search(r"/page/\d+", full_url):
+                pagination_links.append(full_url)
+
+        if not pagination_links:
+            return issues
+
+        # Check rel=prev/next (deprecated by Google but still useful)
+        rel_prev = soup.find("link", attrs={"rel": "prev"})
+        rel_next = soup.find("link", attrs={"rel": "next"})
+        if not rel_prev and not rel_next:
+            issues.append(ProductPageIssue(
+                url=page_url, issue_type="pagination", severity="low",
+                message="No rel=prev/next links on paginated page",
+                recommendation=(
+                    "While Google no longer uses rel=prev/next, other engines may. "
+                    "Consider adding them for broader compatibility."
+                ),
+            ))
+
+        # Check canonical on paginated pages
+        canonical = soup.find("link", attrs={"rel": "canonical"})
+        if canonical:
+            canonical_href = canonical.get("href", "").strip()
+            # If canonical points to page 1 on a non-page-1 URL, flag it
+            parsed_page = urlparse(page_url)
+            page_params = parse_qs(parsed_page.query)
+            current_page_num = page_params.get("page", page_params.get("p", ["1"]))[0]
+
+            if current_page_num != "1":
+                parsed_canonical = urlparse(canonical_href)
+                canon_params = parse_qs(parsed_canonical.query)
+                canon_page_num = canon_params.get("page", canon_params.get("p", ["1"]))[0]
+
+                if canon_page_num == "1" or canonical_href.rstrip("/") == page_url.split("?")[0].rstrip("/"):
+                    issues.append(ProductPageIssue(
+                        url=page_url, issue_type="pagination", severity="high",
+                        message=f"Page {current_page_num} canonical points to page 1",
+                        recommendation=(
+                            "Each paginated page should self-reference its own canonical URL "
+                            "to ensure all pages are indexable."
+                        ),
+                    ))
+
+        # Check robots noindex on filtered/sorted pages
+        robots_meta = soup.find("meta", attrs={"name": "robots"})
+        if robots_meta:
+            content = robots_meta.get("content", "").lower()
+            if "noindex" in content and pagination_links:
+                issues.append(ProductPageIssue(
+                    url=page_url, issue_type="pagination", severity="medium",
+                    message="Paginated page has noindex but contains product links",
+                    recommendation=(
+                        "Ensure products on noindex pages are still discoverable "
+                        "via other indexed pages or sitemap."
+                    ),
+                ))
+
+        return issues
+
+    # ------------------------------------------------------------------
+    # Korean marketplace presence
+    # ------------------------------------------------------------------
+
+    async def check_korean_marketplaces(
+        self,
+        session: aiohttp.ClientSession,
+        brand_name: str,
+    ) -> dict[str, MarketplacePresence]:
+        """Search for brand presence on Korean marketplace platforms."""
+        marketplaces = {}
+
+        search_configs = {
+            "naver_smart_store": {
+                "search_url": f"https://search.shopping.naver.com/search/all?query={brand_name}",
+                "platform": "Naver Smart Store",
+                "indicator_patterns": [r"smartstore\.naver\.com", r"brand\.naver\.com"],
+            },
+            "coupang": {
+                "search_url": f"https://www.coupang.com/np/search?component=&q={brand_name}",
+                "platform": "Coupang",
+                "indicator_patterns": [r"coupang\.com/vp/products/"],
+            },
+            "gmarket": {
+                "search_url": f"https://browse.gmarket.co.kr/search?keyword={brand_name}",
+                "platform": "Gmarket",
+                "indicator_patterns": [r"gmarket\.co\.kr/item/"],
+            },
+            "11st": {
+                "search_url": f"https://search.11st.co.kr/Search.tmall?kwd={brand_name}",
+                "platform": "11번가",
+                "indicator_patterns": [r"11st\.co\.kr/products/"],
+            },
+        }
+
+        for key, cfg in search_configs.items():
+            presence = MarketplacePresence(platform=cfg["platform"], found=False)
+            try:
+                _, html = await self._fetch_page(session, cfg["search_url"])
+                if html:
+                    for pattern in cfg["indicator_patterns"]:
+                        matches = re.findall(pattern, html)
+                        if matches:
+                            presence.found = True
+                            presence.product_count = len(matches)
+                            # Extract first matching URL
+                            url_match = re.search(
+                                rf'href=["\']?(https?://[^"\'>\s]*{pattern}[^"\'>\s]*)',
+                                html,
+                            )
+                            if url_match:
+                                presence.url = url_match.group(1)
+                            break
+            except Exception as exc:
+                self.logger.warning(f"Marketplace check failed for {key}: {exc}")
+
+            marketplaces[key] = presence
+
+        return marketplaces
+
+    # ------------------------------------------------------------------
+    # Naver Smart Store optimization
+    # ------------------------------------------------------------------
+
+    async def check_naver_smart_store(
+        self,
+        session: aiohttp.ClientSession,
+        url: str,
+    ) -> dict[str, Any]:
+        """Check Naver Smart Store-specific SEO elements."""
+        result: dict[str, Any] = {
+            "is_smart_store": False,
+            "issues": [],
+            "optimizations": [],
+        }
+
+        parsed = urlparse(url)
+        is_smart_store = "smartstore.naver.com" in parsed.netloc or "brand.naver.com" in parsed.netloc
+        result["is_smart_store"] = is_smart_store
+
+        _, html = await self._fetch_page(session, url)
+        if not html:
+            return result
+
+        soup = BeautifulSoup(html, "lxml")
+
+        # Check Naver-specific meta tags
+        naver_site_verification = soup.find("meta", attrs={"name": "naver-site-verification"})
+        if not naver_site_verification:
+            result["issues"].append({
+                "type": "naver_verification",
+                "severity": "medium",
+                "message": "Missing naver-site-verification meta tag",
+                "recommendation": "Add Naver Search Advisor verification tag.",
+            })
+
+        # Check for Naver Shopping structured data attributes
+        product_schema = soup.find("script", string=re.compile(r'"@type"\s*:\s*"Product"'))
+        if not product_schema:
+            result["issues"].append({
+                "type": "naver_schema",
+                "severity": "high",
+                "message": "Missing Product schema for Naver Shopping",
+                "recommendation": "Add Product JSON-LD with Korean product names and descriptions.",
+            })
+
+        # Check Korean content optimization
+        body_text = soup.get_text(separator=" ", strip=True)
+        korean_chars = len(re.findall(r"[\uac00-\ud7af]", body_text))
+        total_chars = len(body_text)
+        if total_chars > 0:
+            korean_ratio = korean_chars / total_chars
+            if korean_ratio < 0.3 and is_smart_store:
+                result["issues"].append({
+                    "type": "korean_content",
+                    "severity": "medium",
+                    "message": f"Low Korean content ratio ({korean_ratio:.0%}) for Korean marketplace",
+                    "recommendation": "Increase Korean language content for Naver search visibility.",
+                })
+
+        # Smart Store specific: check product detail image text
+        detail_images = soup.find_all("img", attrs={"class": re.compile(r"detail|product", re.I)})
+        if detail_images and not soup.find("div", attrs={"class": re.compile(r"product.*(desc|detail|content)", re.I)}):
+            result["optimizations"].append({
+                "type": "detail_text",
+                "message": "Product details appear to be image-only",
+                "recommendation": (
+                    "Add HTML text product descriptions alongside images "
+                    "for Naver search indexing."
+                ),
+            })
+
+        return result
+
+    # ------------------------------------------------------------------
+    # Orchestrator
+    # ------------------------------------------------------------------
+
+    async def audit(
+        self,
+        url: str,
+        scope: str = "all",
+        sample: int = 50,
+        check_marketplaces: bool = False,
+    ) -> EcommerceAuditResult:
+        """Run the full e-commerce SEO audit."""
+        result = EcommerceAuditResult(url=url, timestamp=datetime.now().isoformat())
+        parsed = urlparse(url if url.startswith("http") else f"https://{url}")
+        domain = f"{parsed.scheme}://{parsed.netloc}"
+
+        async with aiohttp.ClientSession() as session:
+            # --- Product page audit ---
+            if scope in ("all", "products"):
+                self.logger.info("=== Product Page Audit ===")
+                pages = await self.get_product_pages(domain, sample=sample)
+                result.product_pages_audited = len(pages)
+
+                for page_info in pages:
+                    page_issues = await self.audit_product_page(session, page_info["url"])
+                    for issue in page_issues:
+                        result.add_issue(issue)
+
+                # Duplicate detection
+                all_urls = [p["url"] for p in pages]
+                sitemap_urls = await self._fetch_sitemap_urls(domain)
+                all_urls.extend(sitemap_urls)
+                dup_groups = await self.detect_duplicates(list(set(all_urls)))
+                result.duplicate_groups = [asdict(dg) for dg in dup_groups]
+                for dg in dup_groups:
+                    result.add_issue(ProductPageIssue(
+                        url=dg.canonical_url,
+                        issue_type="duplicate",
+                        severity="high" if dg.reason == "parameter_variant" else "medium",
+                        message=f"Duplicate group ({dg.reason}): {len(dg.duplicate_urls)} variants",
+                        recommendation="Implement canonical tags or parameter handling in GSC/Naver.",
+                    ))
+
+                # Pagination check on category-like pages
+                category_like = [u for u in sitemap_urls if is_category_url(u)][:10]
+                for cat_url in category_like:
+                    pag_issues = await self.check_pagination_seo(session, cat_url)
+                    result.pagination_issues.extend([asdict(i) for i in pag_issues])
+                    for issue in pag_issues:
+                        result.add_issue(issue)
+
+            # --- Category taxonomy ---
+            if scope in ("all", "categories"):
+                self.logger.info("=== Category Taxonomy Analysis ===")
+                cat_result = await self.analyze_category_taxonomy(session, domain)
+                result.category_structure = cat_result
+
+                if cat_result.get("max_depth", 0) > 4:
+                    result.add_issue(ProductPageIssue(
+                        url=domain,
+                        issue_type="category_depth",
+                        severity="medium",
+                        message=f"Category depth exceeds 4 levels (max: {cat_result['max_depth']})",
+                        recommendation="Flatten category structure to 3-4 levels for better crawlability.",
+                    ))
+
+                if cat_result.get("breadcrumbs_missing", 0) > 0:
+                    missing = cat_result["breadcrumbs_missing"]
+                    total = cat_result.get("categories_found", 1)
+                    result.add_issue(ProductPageIssue(
+                        url=domain,
+                        issue_type="breadcrumb",
+                        severity="high" if missing > total * 0.5 else "medium",
+                        message=f"{missing} category pages missing breadcrumb navigation",
+                        recommendation="Add BreadcrumbList schema and visible breadcrumbs to all category pages.",
+                    ))
+
+                for fni in cat_result.get("faceted_nav_issues", []):
+                    result.add_issue(ProductPageIssue(
+                        url=fni["url"],
+                        issue_type="faceted_nav",
+                        severity="high",
+                        message=fni["message"],
+                        recommendation=fni["recommendation"],
+                    ))
+
+            # --- Korean marketplaces ---
+            if check_marketplaces:
+                self.logger.info("=== Korean Marketplace Presence ===")
+                # Extract brand name from site
+                _, home_html = await self._fetch_page(session, domain)
+                brand_name = ""
+                if home_html:
+                    home_soup = BeautifulSoup(home_html, "lxml")
+                    og_site = home_soup.find("meta", attrs={"property": "og:site_name"})
+                    if og_site:
+                        brand_name = og_site.get("content", "").strip()
+                    if not brand_name:
+                        title_tag = home_soup.find("title")
+                        if title_tag:
+                            brand_name = title_tag.get_text(strip=True).split("|")[0].split("-")[0].strip()
+
+                if brand_name:
+                    mp_results = await self.check_korean_marketplaces(session, brand_name)
+                    result.korean_marketplaces = {
+                        k: asdict(v) for k, v in mp_results.items()
+                    }
+
+                # Naver Smart Store check
+                naver_result = await self.check_naver_smart_store(session, domain)
+                result.naver_smart_store = naver_result
+                for naver_issue in naver_result.get("issues", []):
+                    result.add_issue(ProductPageIssue(
+                        url=domain,
+                        issue_type=naver_issue["type"],
+                        severity=naver_issue["severity"],
+                        message=naver_issue["message"],
+                        recommendation=naver_issue["recommendation"],
+                    ))
+
+        result.calculate_score()
+        return result
+
+
+# ---------------------------------------------------------------------------
+# CLI output helpers
+# ---------------------------------------------------------------------------
+
+def print_rich_report(result: EcommerceAuditResult) -> None:
+    """Print a rich-formatted report to the console."""
+    console.print(f"\n[bold cyan]E-Commerce SEO Audit Report[/bold cyan]")
+    console.print(f"URL: {result.url}")
+    console.print(f"Product Pages Audited: {result.product_pages_audited}")
+    console.print(f"Timestamp: {result.timestamp}")
+
+    # Score
+    score_color = "green" if result.score >= 80 else "yellow" if result.score >= 50 else "red"
+    console.print(f"\n[bold {score_color}]Score: {result.score}/100[/bold {score_color}]")
+
+    # Issues summary
+    table = Table(title="Issues Summary")
+    table.add_column("Severity", style="bold")
+    table.add_column("Count", justify="right")
+    for sev in ["critical", "high", "medium", "low"]:
+        color = {"critical": "red", "high": "yellow", "medium": "cyan", "low": "dim"}[sev]
+        table.add_row(f"[{color}]{sev.upper()}[/{color}]", str(len(result.issues[sev])))
+    console.print(table)
+
+    # Top issues
+    for sev in ["critical", "high"]:
+        if result.issues[sev]:
+            console.print(f"\n[bold red]{sev.upper()} Issues:[/bold red]")
+            for issue in result.issues[sev][:10]:
+                console.print(f"  - [{issue['issue_type']}] {issue['message']}")
+                console.print(f"    [dim]{issue['recommendation']}[/dim]")
+
+    # Category structure
+    if result.category_structure:
+        cs = result.category_structure
+        console.print(f"\n[bold]Category Structure:[/bold]")
+        console.print(f"  Categories found: {cs.get('categories_found', 0)}")
+        console.print(f"  Max depth: {cs.get('max_depth', 0)}")
+        console.print(f"  Breadcrumbs present: {cs.get('breadcrumbs_present', 0)}")
+        console.print(f"  Breadcrumbs missing: {cs.get('breadcrumbs_missing', 0)}")
+
+    # Duplicates
+    if result.duplicate_groups:
+        console.print(f"\n[bold]Duplicate Groups: {len(result.duplicate_groups)}[/bold]")
+        for dg in result.duplicate_groups[:5]:
+            console.print(f"  [{dg['reason']}] {dg['canonical_url']} ({len(dg['duplicate_urls'])} variants)")
+
+    # Korean marketplaces
+    if result.korean_marketplaces:
+        console.print(f"\n[bold]Korean Marketplace Presence:[/bold]")
+        for key, mp in result.korean_marketplaces.items():
+            status = "[green]Found[/green]" if mp.get("found") else "[red]Not Found[/red]"
+            console.print(f"  {mp.get('platform', key)}: {status}")
+            if mp.get("url"):
+                console.print(f"    URL: {mp['url']}")
+
+
+# ---------------------------------------------------------------------------
+# Main
+# ---------------------------------------------------------------------------
+
+def main() -> None:
+    parser = argparse.ArgumentParser(
+        description="E-Commerce SEO Auditor - Product page and marketplace audit",
+    )
+    parser.add_argument("--url", required=True, help="Target website URL")
+    parser.add_argument(
+        "--scope",
+        choices=["all", "products", "categories"],
+        default="all",
+        help="Audit scope (default: all)",
+    )
+    parser.add_argument(
+        "--korean-marketplaces",
+        action="store_true",
+        help="Check Korean marketplace presence (Coupang, Gmarket, 11번가, Naver)",
+    )
+    parser.add_argument(
+        "--sample",
+        type=int,
+        default=50,
+        help="Number of product pages to sample (default: 50)",
+    )
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", type=str, help="Save output to file")
+    args = parser.parse_args()
+
+    auditor = EcommerceAuditor()
+    result = asyncio.run(
+        auditor.audit(
+            url=args.url,
+            scope=args.scope,
+            sample=args.sample,
+            check_marketplaces=args.korean_marketplaces,
+        )
+    )
+
+    if args.json:
+        output = json.dumps(asdict(result), indent=2, ensure_ascii=False, default=str)
+        if args.output:
+            with open(args.output, "w", encoding="utf-8") as f:
+                f.write(output)
+            console.print(f"[green]Results saved to {args.output}[/green]")
+        else:
+            print(output)
+    else:
+        print_rich_report(result)
+        if args.output:
+            output = json.dumps(asdict(result), indent=2, ensure_ascii=False, default=str)
+            with open(args.output, "w", encoding="utf-8") as f:
+                f.write(output)
+            console.print(f"\n[green]JSON results also saved to {args.output}[/green]")
+
+    auditor.print_stats()
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/24-seo-ecommerce/code/scripts/product_schema_checker.py b/custom-skills/24-seo-ecommerce/code/scripts/product_schema_checker.py
new file mode 100644
index 0000000..dad07e2
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/code/scripts/product_schema_checker.py
@@ -0,0 +1,805 @@
+"""
+Product Schema Checker
+======================
+Purpose: Validate Product structured data (JSON-LD, Microdata, RDFa)
+         for Google and Naver rich result eligibility.
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import re
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from typing import Any
+from urllib.parse import urljoin, urlparse
+
+import aiohttp
+from bs4 import BeautifulSoup
+from rich.console import Console
+from rich.table import Table
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+console = Console()
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class SchemaProperty:
+    """Single property within a schema object."""
+    name: str
+    value: Any
+    required: bool
+    valid: bool
+    error: str = ""
+
+
+@dataclass
+class ProductSchema:
+    """Validation result for one product schema on a page."""
+    url: str
+    schema_type: str               # Product, Offer, AggregateRating, etc.
+    properties: list[dict]          # list of SchemaProperty as dicts
+    is_valid: bool = False
+    rich_result_eligible: bool = False
+    errors: list[str] = field(default_factory=list)
+    warnings: list[str] = field(default_factory=list)
+
+
+@dataclass
+class SchemaCheckResult:
+    """Complete schema check result for one or more pages."""
+    urls_checked: int = 0
+    pages_with_schema: int = 0
+    pages_without_schema: int = 0
+    schemas: list[dict] = field(default_factory=list)
+    common_errors: list[str] = field(default_factory=list)
+    common_warnings: list[str] = field(default_factory=list)
+    naver_shopping_issues: list[dict] = field(default_factory=list)
+    score: int = 0
+    timestamp: str = ""
+
+    def calculate_score(self) -> int:
+        """Score 0-100 based on schema completeness."""
+        if self.urls_checked == 0:
+            self.score = 0
+            return 0
+        coverage = self.pages_with_schema / self.urls_checked
+        valid_schemas = sum(1 for s in self.schemas if s.get("is_valid"))
+        validity_rate = valid_schemas / max(len(self.schemas), 1)
+        eligible = sum(1 for s in self.schemas if s.get("rich_result_eligible"))
+        eligibility_rate = eligible / max(len(self.schemas), 1)
+        self.score = int(coverage * 40 + validity_rate * 35 + eligibility_rate * 25)
+        return self.score
+
+
+# ---------------------------------------------------------------------------
+# Schema requirements
+# ---------------------------------------------------------------------------
+
+PRODUCT_REQUIRED = {"name", "image", "description"}
+PRODUCT_RECOMMENDED = {
+    "brand", "sku", "gtin", "gtin8", "gtin13", "gtin14", "mpn",
+    "offers", "review", "aggregateRating", "color", "material",
+}
+
+OFFER_REQUIRED = {"price", "priceCurrency", "availability"}
+OFFER_RECOMMENDED = {
+    "url", "priceValidUntil", "itemCondition", "seller",
+    "shippingDetails", "hasMerchantReturnPolicy",
+}
+
+AGGREGATE_RATING_REQUIRED = {"ratingValue", "reviewCount"}
+AGGREGATE_RATING_RECOMMENDED = {"bestRating", "worstRating", "ratingCount"}
+
+REVIEW_REQUIRED = {"author", "reviewRating"}
+REVIEW_RECOMMENDED = {"datePublished", "reviewBody", "name"}
+
+BREADCRUMB_REQUIRED = {"itemListElement"}
+
+AVAILABILITY_VALUES = {
+    "https://schema.org/InStock",
+    "https://schema.org/OutOfStock",
+    "https://schema.org/PreOrder",
+    "https://schema.org/BackOrder",
+    "https://schema.org/Discontinued",
+    "https://schema.org/InStoreOnly",
+    "https://schema.org/OnlineOnly",
+    "https://schema.org/LimitedAvailability",
+    "https://schema.org/SoldOut",
+    "http://schema.org/InStock",
+    "http://schema.org/OutOfStock",
+    "http://schema.org/PreOrder",
+    "http://schema.org/BackOrder",
+    "http://schema.org/Discontinued",
+    "InStock", "OutOfStock", "PreOrder", "BackOrder", "Discontinued",
+}
+
+ITEM_CONDITION_VALUES = {
+    "https://schema.org/NewCondition",
+    "https://schema.org/UsedCondition",
+    "https://schema.org/RefurbishedCondition",
+    "https://schema.org/DamagedCondition",
+    "http://schema.org/NewCondition",
+    "http://schema.org/UsedCondition",
+    "http://schema.org/RefurbishedCondition",
+    "NewCondition", "UsedCondition", "RefurbishedCondition",
+}
+
+
+# ---------------------------------------------------------------------------
+# Main checker
+# ---------------------------------------------------------------------------
+
+class ProductSchemaChecker(BaseAsyncClient):
+    """Validate Product structured data on e-commerce pages."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 10,
+        requests_per_second: float = 5.0,
+        timeout: int = 30,
+    ):
+        super().__init__(max_concurrent=max_concurrent, requests_per_second=requests_per_second)
+        self.timeout = aiohttp.ClientTimeout(total=timeout)
+        self.headers = {
+            "User-Agent": (
+                "Mozilla/5.0 (compatible; ProductSchemaChecker/1.0; "
+                "+https://ourdigital.org)"
+            ),
+            "Accept": "text/html,application/xhtml+xml",
+            "Accept-Language": "ko-KR,ko;q=0.9,en-US;q=0.8,en;q=0.7",
+        }
+
+    # ------------------------------------------------------------------
+    # Page fetching
+    # ------------------------------------------------------------------
+
+    async def _fetch_page(self, session: aiohttp.ClientSession, url: str) -> str:
+        """Fetch page HTML."""
+        try:
+            async with session.get(url, headers=self.headers, timeout=self.timeout,
+                                   allow_redirects=True, ssl=False) as resp:
+                return await resp.text(errors="replace")
+        except Exception as exc:
+            self.logger.warning(f"Failed to fetch {url}: {exc}")
+            return ""
+
+    # ------------------------------------------------------------------
+    # Schema extraction
+    # ------------------------------------------------------------------
+
+    def extract_schemas(self, html: str, page_url: str) -> list[dict]:
+        """Extract all structured data from HTML (JSON-LD, Microdata, RDFa)."""
+        schemas: list[dict] = []
+        soup = BeautifulSoup(html, "lxml")
+
+        # --- JSON-LD ---
+        for script in soup.find_all("script", attrs={"type": "application/ld+json"}):
+            try:
+                text = script.string or script.get_text()
+                if not text:
+                    continue
+                data = json.loads(text)
+                if isinstance(data, list):
+                    for item in data:
+                        if isinstance(item, dict):
+                            schemas.append(item)
+                elif isinstance(data, dict):
+                    # Handle @graph
+                    if "@graph" in data:
+                        for item in data["@graph"]:
+                            if isinstance(item, dict):
+                                schemas.append(item)
+                    else:
+                        schemas.append(data)
+            except (json.JSONDecodeError, TypeError) as exc:
+                self.logger.debug(f"JSON-LD parse error on {page_url}: {exc}")
+
+        # --- Microdata ---
+        for item_scope in soup.find_all(attrs={"itemscope": True}):
+            item_type = item_scope.get("itemtype", "")
+            if "Product" in item_type or "Offer" in item_type:
+                microdata = self._parse_microdata(item_scope)
+                if microdata:
+                    schemas.append(microdata)
+
+        return schemas
+
+    def _parse_microdata(self, element) -> dict:
+        """Parse microdata from an itemscope element."""
+        result: dict[str, Any] = {}
+        item_type = element.get("itemtype", "")
+        if item_type:
+            type_name = item_type.rstrip("/").split("/")[-1]
+            result["@type"] = type_name
+
+        for prop in element.find_all(attrs={"itemprop": True}, recursive=True):
+            name = prop.get("itemprop", "")
+            if not name:
+                continue
+            # Nested itemscope
+            if prop.get("itemscope") is not None:
+                result[name] = self._parse_microdata(prop)
+            elif prop.name == "meta":
+                result[name] = prop.get("content", "")
+            elif prop.name == "link":
+                result[name] = prop.get("href", "")
+            elif prop.name == "img":
+                result[name] = prop.get("src", "")
+            elif prop.name == "time":
+                result[name] = prop.get("datetime", prop.get_text(strip=True))
+            else:
+                result[name] = prop.get_text(strip=True)
+
+        return result
+
+    # ------------------------------------------------------------------
+    # Validation methods
+    # ------------------------------------------------------------------
+
+    def validate_product_schema(self, schema_data: dict, page_url: str) -> ProductSchema:
+        """Validate a Product schema object."""
+        ps = ProductSchema(
+            url=page_url,
+            schema_type="Product",
+            properties=[],
+        )
+
+        # Check required properties
+        for prop_name in PRODUCT_REQUIRED:
+            value = schema_data.get(prop_name)
+            valid = bool(value)
+            error = "" if valid else f"Missing required property: {prop_name}"
+            sp = SchemaProperty(
+                name=prop_name, value=value, required=True, valid=valid, error=error,
+            )
+            ps.properties.append(asdict(sp))
+            if not valid:
+                ps.errors.append(error)
+
+        # Check recommended properties
+        for prop_name in PRODUCT_RECOMMENDED:
+            value = schema_data.get(prop_name)
+            sp = SchemaProperty(
+                name=prop_name, value=value if value else None,
+                required=False, valid=bool(value),
+                error="" if value else f"Missing recommended property: {prop_name}",
+            )
+            ps.properties.append(asdict(sp))
+            if not value:
+                ps.warnings.append(f"Missing recommended property: {prop_name}")
+
+        # Validate offers
+        offers = schema_data.get("offers")
+        if offers:
+            if isinstance(offers, list):
+                for offer in offers:
+                    offer_errors = self.validate_offer_schema(offer)
+                    ps.errors.extend(offer_errors["errors"])
+                    ps.warnings.extend(offer_errors["warnings"])
+            elif isinstance(offers, dict):
+                offer_errors = self.validate_offer_schema(offers)
+                ps.errors.extend(offer_errors["errors"])
+                ps.warnings.extend(offer_errors["warnings"])
+        else:
+            ps.errors.append("Missing 'offers' property (required for rich results)")
+
+        # Validate aggregateRating
+        agg_rating = schema_data.get("aggregateRating")
+        if agg_rating and isinstance(agg_rating, dict):
+            rating_result = self.validate_aggregate_rating(agg_rating)
+            ps.errors.extend(rating_result["errors"])
+            ps.warnings.extend(rating_result["warnings"])
+
+        # Validate reviews
+        review = schema_data.get("review")
+        if review:
+            reviews = review if isinstance(review, list) else [review]
+            for r in reviews[:5]:  # Check up to 5 reviews
+                if isinstance(r, dict):
+                    review_result = self.validate_review_schema(r)
+                    ps.errors.extend(review_result["errors"])
+                    ps.warnings.extend(review_result["warnings"])
+
+        ps.is_valid = len(ps.errors) == 0
+        ps.rich_result_eligible = self.check_rich_result_eligibility(schema_data)
+
+        return ps
+
+    def validate_offer_schema(self, offer_data: dict) -> dict[str, list[str]]:
+        """Validate an Offer schema object."""
+        errors: list[str] = []
+        warnings: list[str] = []
+
+        for prop_name in OFFER_REQUIRED:
+            value = offer_data.get(prop_name)
+            if not value:
+                errors.append(f"Offer missing required property: {prop_name}")
+
+        # Validate price format
+        price = offer_data.get("price")
+        if price is not None:
+            price_str = str(price).replace(",", "").strip()
+            if not re.match(r"^\d+(\.\d+)?$", price_str):
+                errors.append(f"Invalid price format: '{price}' (must be numeric)")
+            elif float(price_str) <= 0:
+                warnings.append(f"Price is zero or negative: {price}")
+
+        # Validate priceCurrency
+        currency = offer_data.get("priceCurrency", "")
+        valid_currencies = {"KRW", "USD", "EUR", "JPY", "CNY", "GBP"}
+        if currency and currency.upper() not in valid_currencies:
+            warnings.append(f"Unusual currency code: {currency}")
+
+        # Validate availability
+        availability = offer_data.get("availability", "")
+        if availability and availability not in AVAILABILITY_VALUES:
+            errors.append(
+                f"Invalid availability value: '{availability}'. "
+                f"Use schema.org values like https://schema.org/InStock"
+            )
+
+        # Validate itemCondition
+        condition = offer_data.get("itemCondition", "")
+        if condition and condition not in ITEM_CONDITION_VALUES:
+            warnings.append(f"Invalid itemCondition: '{condition}'")
+
+        # Check recommended
+        for prop_name in OFFER_RECOMMENDED:
+            if not offer_data.get(prop_name):
+                warnings.append(f"Offer missing recommended property: {prop_name}")
+
+        return {"errors": errors, "warnings": warnings}
+
+    def validate_aggregate_rating(self, rating_data: dict) -> dict[str, list[str]]:
+        """Validate AggregateRating schema."""
+        errors: list[str] = []
+        warnings: list[str] = []
+
+        for prop_name in AGGREGATE_RATING_REQUIRED:
+            value = rating_data.get(prop_name)
+            if value is None:
+                errors.append(f"AggregateRating missing required: {prop_name}")
+
+        # Validate ratingValue range
+        rating_value = rating_data.get("ratingValue")
+        best_rating = rating_data.get("bestRating", 5)
+        worst_rating = rating_data.get("worstRating", 1)
+        if rating_value is not None:
+            try:
+                rv = float(rating_value)
+                br = float(best_rating)
+                wr = float(worst_rating)
+                if rv < wr or rv > br:
+                    errors.append(
+                        f"ratingValue ({rv}) outside range [{wr}, {br}]"
+                    )
+            except (ValueError, TypeError):
+                errors.append(f"Invalid ratingValue format: {rating_value}")
+
+        # Validate reviewCount
+        review_count = rating_data.get("reviewCount")
+        if review_count is not None:
+            try:
+                rc = int(review_count)
+                if rc < 0:
+                    errors.append(f"Negative reviewCount: {rc}")
+            except (ValueError, TypeError):
+                errors.append(f"Invalid reviewCount format: {review_count}")
+
+        for prop_name in AGGREGATE_RATING_RECOMMENDED:
+            if not rating_data.get(prop_name):
+                warnings.append(f"AggregateRating missing recommended: {prop_name}")
+
+        return {"errors": errors, "warnings": warnings}
+
+    def validate_review_schema(self, review_data: dict) -> dict[str, list[str]]:
+        """Validate Review schema."""
+        errors: list[str] = []
+        warnings: list[str] = []
+
+        # Author validation
+        author = review_data.get("author")
+        if not author:
+            errors.append("Review missing required: author")
+        elif isinstance(author, dict):
+            author_name = author.get("name", "")
+            if not author_name:
+                errors.append("Review author missing 'name' property")
+        elif isinstance(author, str):
+            if len(author.strip()) == 0:
+                errors.append("Review author is empty string")
+
+        # reviewRating validation
+        review_rating = review_data.get("reviewRating")
+        if not review_rating:
+            errors.append("Review missing required: reviewRating")
+        elif isinstance(review_rating, dict):
+            rv = review_rating.get("ratingValue")
+            if rv is None:
+                errors.append("reviewRating missing ratingValue")
+
+        for prop_name in REVIEW_RECOMMENDED:
+            if not review_data.get(prop_name):
+                warnings.append(f"Review missing recommended: {prop_name}")
+
+        return {"errors": errors, "warnings": warnings}
+
+    def validate_breadcrumb(self, schema_data: dict) -> dict[str, list[str]]:
+        """Validate BreadcrumbList schema."""
+        errors: list[str] = []
+        warnings: list[str] = []
+
+        items = schema_data.get("itemListElement")
+        if not items:
+            errors.append("BreadcrumbList missing itemListElement")
+            return {"errors": errors, "warnings": warnings}
+
+        if not isinstance(items, list):
+            errors.append("itemListElement should be an array")
+            return {"errors": errors, "warnings": warnings}
+
+        for i, item in enumerate(items):
+            if not isinstance(item, dict):
+                errors.append(f"Breadcrumb item {i} is not an object")
+                continue
+            position = item.get("position")
+            if position is None:
+                errors.append(f"Breadcrumb item {i} missing 'position'")
+            name = item.get("name") or (item.get("item", {}).get("name") if isinstance(item.get("item"), dict) else None)
+            if not name:
+                warnings.append(f"Breadcrumb item {i} missing 'name'")
+
+        return {"errors": errors, "warnings": warnings}
+
+    # ------------------------------------------------------------------
+    # Rich result eligibility
+    # ------------------------------------------------------------------
+
+    def check_rich_result_eligibility(self, schema_data: dict) -> bool:
+        """Assess Google rich result eligibility for Product schema."""
+        # Must have name, image, and offers with price
+        if not schema_data.get("name"):
+            return False
+        if not schema_data.get("image"):
+            return False
+
+        offers = schema_data.get("offers")
+        if not offers:
+            return False
+
+        offer_list = offers if isinstance(offers, list) else [offers]
+        for offer in offer_list:
+            if not isinstance(offer, dict):
+                continue
+            if offer.get("price") and offer.get("priceCurrency") and offer.get("availability"):
+                return True
+
+        return False
+
+    # ------------------------------------------------------------------
+    # Naver Shopping requirements
+    # ------------------------------------------------------------------
+
+    def check_naver_shopping_requirements(self, schema_data: dict, page_url: str) -> list[dict]:
+        """Check Naver Shopping specific schema requirements."""
+        issues: list[dict] = []
+
+        # Naver Shopping requires Product name in Korean for Korean market
+        name = schema_data.get("name", "")
+        korean_chars = len(re.findall(r"[\uac00-\ud7af]", str(name)))
+        if korean_chars == 0 and name:
+            issues.append({
+                "url": page_url,
+                "type": "naver_product_name",
+                "severity": "medium",
+                "message": "Product name has no Korean characters",
+                "recommendation": "Include Korean product name for Naver Shopping visibility.",
+            })
+
+        # Naver prefers specific category mapping
+        if not schema_data.get("category"):
+            issues.append({
+                "url": page_url,
+                "type": "naver_category",
+                "severity": "low",
+                "message": "Missing 'category' property for Naver Shopping categorization",
+                "recommendation": "Add category property matching Naver Shopping category taxonomy.",
+            })
+
+        # Naver requires image
+        image = schema_data.get("image")
+        if not image:
+            issues.append({
+                "url": page_url,
+                "type": "naver_image",
+                "severity": "high",
+                "message": "Missing product image (required for Naver Shopping)",
+                "recommendation": "Add at least one high-quality product image URL.",
+            })
+        elif isinstance(image, str):
+            if not image.startswith("http"):
+                issues.append({
+                    "url": page_url,
+                    "type": "naver_image_url",
+                    "severity": "medium",
+                    "message": "Product image URL is relative (should be absolute)",
+                    "recommendation": "Use absolute URLs for product images.",
+                })
+
+        # Naver requires price in KRW
+        offers = schema_data.get("offers")
+        if offers:
+            offer_list = offers if isinstance(offers, list) else [offers]
+            for offer in offer_list:
+                if isinstance(offer, dict):
+                    currency = offer.get("priceCurrency", "")
+                    if currency and currency.upper() != "KRW":
+                        issues.append({
+                            "url": page_url,
+                            "type": "naver_currency",
+                            "severity": "medium",
+                            "message": f"Price currency is {currency}, not KRW",
+                            "recommendation": "For Naver Shopping, provide price in KRW.",
+                        })
+
+        # Check brand/manufacturer
+        if not schema_data.get("brand") and not schema_data.get("manufacturer"):
+            issues.append({
+                "url": page_url,
+                "type": "naver_brand",
+                "severity": "low",
+                "message": "Missing brand/manufacturer (helpful for Naver Shopping filters)",
+                "recommendation": "Add brand or manufacturer property.",
+            })
+
+        return issues
+
+    # ------------------------------------------------------------------
+    # Orchestrator
+    # ------------------------------------------------------------------
+
+    async def check(
+        self,
+        urls: list[str] | None = None,
+        sitemap_url: str | None = None,
+        sample_size: int = 50,
+    ) -> SchemaCheckResult:
+        """Run schema validation on URLs or sitemap."""
+        result = SchemaCheckResult(timestamp=datetime.now().isoformat())
+        target_urls: list[str] = []
+
+        async with aiohttp.ClientSession() as session:
+            if sitemap_url:
+                # Fetch URLs from sitemap
+                target_urls = await self._urls_from_sitemap(session, sitemap_url, sample_size)
+            if urls:
+                target_urls.extend(urls)
+
+            target_urls = list(set(target_urls))[:sample_size]
+            result.urls_checked = len(target_urls)
+            self.logger.info(f"Checking {len(target_urls)} URLs for Product schema")
+
+            error_counter: dict[str, int] = {}
+            warning_counter: dict[str, int] = {}
+
+            for url in target_urls:
+                html = await self._fetch_page(session, url)
+                if not html:
+                    result.pages_without_schema += 1
+                    continue
+
+                schemas = self.extract_schemas(html, url)
+                product_schemas = [
+                    s for s in schemas
+                    if self._get_schema_type(s) in ("Product", "ProductGroup")
+                ]
+                breadcrumb_schemas = [
+                    s for s in schemas
+                    if self._get_schema_type(s) == "BreadcrumbList"
+                ]
+
+                if not product_schemas:
+                    result.pages_without_schema += 1
+                    continue
+
+                result.pages_with_schema += 1
+
+                for ps_data in product_schemas:
+                    ps = self.validate_product_schema(ps_data, url)
+                    result.schemas.append(asdict(ps))
+
+                    for err in ps.errors:
+                        error_counter[err] = error_counter.get(err, 0) + 1
+                    for warn in ps.warnings:
+                        warning_counter[warn] = warning_counter.get(warn, 0) + 1
+
+                    # Naver Shopping checks
+                    naver_issues = self.check_naver_shopping_requirements(ps_data, url)
+                    result.naver_shopping_issues.extend(naver_issues)
+
+                # Validate breadcrumbs
+                for bc_data in breadcrumb_schemas:
+                    bc_result = self.validate_breadcrumb(bc_data)
+                    for err in bc_result["errors"]:
+                        error_counter[err] = error_counter.get(err, 0) + 1
+
+            # Aggregate common errors/warnings
+            result.common_errors = sorted(
+                error_counter.keys(),
+                key=lambda k: error_counter[k],
+                reverse=True,
+            )[:20]
+            result.common_warnings = sorted(
+                warning_counter.keys(),
+                key=lambda k: warning_counter[k],
+                reverse=True,
+            )[:20]
+
+        result.calculate_score()
+        return result
+
+    async def _urls_from_sitemap(
+        self,
+        session: aiohttp.ClientSession,
+        sitemap_url: str,
+        limit: int,
+    ) -> list[str]:
+        """Fetch product URLs from sitemap."""
+        urls: list[str] = []
+        try:
+            async with session.get(sitemap_url, headers=self.headers,
+                                   timeout=self.timeout, ssl=False) as resp:
+                if resp.status != 200:
+                    return urls
+                text = await resp.text(errors="replace")
+                soup = BeautifulSoup(text, "lxml-xml")
+
+                # Handle sitemap index
+                sitemapindex = soup.find_all("sitemap")
+                if sitemapindex:
+                    for sm in sitemapindex[:3]:
+                        loc = sm.find("loc")
+                        if loc:
+                            child_urls = await self._urls_from_sitemap(session, loc.text.strip(), limit)
+                            urls.extend(child_urls)
+                            if len(urls) >= limit:
+                                break
+                else:
+                    for tag in soup.find_all("url"):
+                        loc = tag.find("loc")
+                        if loc:
+                            urls.append(loc.text.strip())
+                            if len(urls) >= limit:
+                                break
+        except Exception as exc:
+            self.logger.warning(f"Sitemap parse failed: {exc}")
+
+        return urls[:limit]
+
+    @staticmethod
+    def _get_schema_type(schema: dict) -> str:
+        """Get the @type from a schema dict, handling various formats."""
+        schema_type = schema.get("@type", "")
+        if isinstance(schema_type, list):
+            return schema_type[0] if schema_type else ""
+        return str(schema_type)
+
+
+# ---------------------------------------------------------------------------
+# CLI output helpers
+# ---------------------------------------------------------------------------
+
+def print_rich_report(result: SchemaCheckResult) -> None:
+    """Print a rich-formatted report to the console."""
+    console.print(f"\n[bold cyan]Product Schema Validation Report[/bold cyan]")
+    console.print(f"Timestamp: {result.timestamp}")
+    console.print(f"URLs checked: {result.urls_checked}")
+
+    # Coverage
+    coverage = (result.pages_with_schema / max(result.urls_checked, 1)) * 100
+    cov_color = "green" if coverage >= 90 else "yellow" if coverage >= 50 else "red"
+    console.print(f"Schema coverage: [{cov_color}]{coverage:.0f}%[/{cov_color}] "
+                  f"({result.pages_with_schema}/{result.urls_checked})")
+
+    # Score
+    score_color = "green" if result.score >= 80 else "yellow" if result.score >= 50 else "red"
+    console.print(f"[bold {score_color}]Score: {result.score}/100[/bold {score_color}]")
+
+    # Validity summary
+    valid = sum(1 for s in result.schemas if s.get("is_valid"))
+    eligible = sum(1 for s in result.schemas if s.get("rich_result_eligible"))
+    total = len(result.schemas)
+
+    table = Table(title="Schema Summary")
+    table.add_column("Metric", style="bold")
+    table.add_column("Value", justify="right")
+    table.add_row("Total schemas found", str(total))
+    table.add_row("Valid schemas", str(valid))
+    table.add_row("Rich result eligible", str(eligible))
+    table.add_row("Pages without schema", str(result.pages_without_schema))
+    console.print(table)
+
+    # Common errors
+    if result.common_errors:
+        console.print(f"\n[bold red]Common Errors ({len(result.common_errors)}):[/bold red]")
+        for err in result.common_errors[:10]:
+            console.print(f"  [red]-[/red] {err}")
+
+    # Common warnings
+    if result.common_warnings:
+        console.print(f"\n[bold yellow]Common Warnings ({len(result.common_warnings)}):[/bold yellow]")
+        for warn in result.common_warnings[:10]:
+            console.print(f"  [yellow]-[/yellow] {warn}")
+
+    # Naver Shopping issues
+    if result.naver_shopping_issues:
+        console.print(f"\n[bold magenta]Naver Shopping Issues ({len(result.naver_shopping_issues)}):[/bold magenta]")
+        seen: set[str] = set()
+        for issue in result.naver_shopping_issues:
+            key = f"{issue['type']}:{issue['message']}"
+            if key not in seen:
+                seen.add(key)
+                console.print(f"  [{issue.get('severity', 'medium')}] {issue['message']}")
+                console.print(f"    [dim]{issue['recommendation']}[/dim]")
+
+
+# ---------------------------------------------------------------------------
+# Main
+# ---------------------------------------------------------------------------
+
+def main() -> None:
+    parser = argparse.ArgumentParser(
+        description="Product Schema Checker - Validate e-commerce structured data",
+    )
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument("--url", nargs="+", help="Product page URL(s) to validate")
+    group.add_argument("--sitemap", help="Sitemap URL to fetch product pages from")
+    parser.add_argument(
+        "--sample",
+        type=int,
+        default=50,
+        help="Max URLs to check from sitemap (default: 50)",
+    )
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", type=str, help="Save output to file")
+    args = parser.parse_args()
+
+    checker = ProductSchemaChecker()
+    result = asyncio.run(
+        checker.check(
+            urls=args.url,
+            sitemap_url=args.sitemap,
+            sample_size=args.sample,
+        )
+    )
+
+    if args.json:
+        output = json.dumps(asdict(result), indent=2, ensure_ascii=False, default=str)
+        if args.output:
+            with open(args.output, "w", encoding="utf-8") as f:
+                f.write(output)
+            console.print(f"[green]Results saved to {args.output}[/green]")
+        else:
+            print(output)
+    else:
+        print_rich_report(result)
+        if args.output:
+            output = json.dumps(asdict(result), indent=2, ensure_ascii=False, default=str)
+            with open(args.output, "w", encoding="utf-8") as f:
+                f.write(output)
+            console.print(f"\n[green]JSON results also saved to {args.output}[/green]")
+
+    checker.print_stats()
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/24-seo-ecommerce/code/scripts/requirements.txt b/custom-skills/24-seo-ecommerce/code/scripts/requirements.txt
new file mode 100644
index 0000000..55ae17f
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/code/scripts/requirements.txt
@@ -0,0 +1,9 @@
+# 24-seo-ecommerce dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+beautifulsoup4>=4.12.0
+lxml>=5.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/24-seo-ecommerce/desktop/SKILL.md b/custom-skills/24-seo-ecommerce/desktop/SKILL.md
new file mode 100644
index 0000000..f40db32
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/desktop/SKILL.md
@@ -0,0 +1,156 @@
+---
+name: seo-ecommerce
+description: |
+  E-commerce SEO audit and optimization for product pages, product schema, category taxonomy,
+  and Korean marketplace presence.
+  Triggers: product SEO, e-commerce audit, product schema, category SEO, Smart Store, marketplace SEO,
+  상품 SEO, 이커머스 감사, 쇼핑몰 SEO.
+---
+
+# E-Commerce SEO Audit
+
+## Purpose
+
+Audit e-commerce sites for product page optimization, structured data validation, category taxonomy health, duplicate content issues, and Korean marketplace presence (Naver Smart Store, Coupang, Gmarket, 11번가).
+
+## Core Capabilities
+
+1. **Product Page SEO Audit** - Title, meta description, H1, image alt text, internal links, canonical tags
+2. **Product Schema Validation** - Product, Offer, AggregateRating, Review, BreadcrumbList structured data
+3. **Category Taxonomy Analysis** - Depth check, breadcrumbs, faceted navigation handling
+4. **Duplicate Content Detection** - Parameter variants, product variants, pagination issues
+5. **Korean Marketplace Presence** - Naver Smart Store, Coupang, Gmarket, 11번가
+
+## MCP Tool Usage
+
+### Ahrefs for Product Page Discovery
+```
+mcp__ahrefs__site-explorer-pages-by-traffic: Identify top product and category pages
+mcp__ahrefs__site-explorer-organic-keywords: Product page keyword performance
+```
+
+### WebSearch for Marketplace Checks
+```
+WebSearch: Search for brand presence on Korean marketplaces
+WebFetch: Fetch and analyze marketplace listing pages
+```
+
+### Notion for Report Storage
+```
+mcp__notion__notion-create-pages: Save audit report to SEO Audit Log database
+```
+
+## Workflow
+
+### 1. Product Page Audit
+1. Discover product pages via Ahrefs pages-by-traffic or sitemap
+2. For each product page check:
+   - Title tag: contains product name, under 60 chars
+   - Meta description: includes price/feature info, under 155 chars
+   - Single H1 with product name
+   - All product images have descriptive alt text
+   - Canonical tag present and correct
+   - Sufficient internal links (related products, breadcrumbs)
+   - Open Graph tags for social sharing
+3. Score severity: critical/high/medium/low
+
+### 2. Product Schema Validation
+1. Extract JSON-LD and Microdata from product pages
+2. Validate Product type: name, image, description (required)
+3. Validate Offer: price, priceCurrency, availability (required)
+4. Validate AggregateRating: ratingValue, reviewCount (required)
+5. Validate Review: author, reviewRating (required)
+6. Check BreadcrumbList implementation
+7. Assess Google rich result eligibility
+8. Check Naver Shopping specific requirements (Korean name, KRW price, absolute image URLs)
+
+### 3. Category Taxonomy Analysis
+1. Crawl category pages from sitemap or homepage navigation
+2. Measure taxonomy depth (warn if > 4 levels)
+3. Check breadcrumb presence on every category page
+4. Identify faceted navigation URLs that are indexable without proper canonicals
+5. Count child category links for structure assessment
+
+### 4. Duplicate Content Detection
+1. Group URLs by base path (stripping query parameters)
+2. Identify parameter variants (?color=, ?size=, ?sort=)
+3. Detect product variant URL duplicates (e.g., /product-red vs /product-blue)
+4. Flag paginated pages missing self-referencing canonicals
+
+### 5. Korean Marketplace Presence
+1. Extract brand name from site (og:site_name or title)
+2. Search each marketplace for brand products:
+   - Naver Smart Store (smartstore.naver.com)
+   - Coupang (coupang.com)
+   - Gmarket (gmarket.co.kr)
+   - 11번가 (11st.co.kr)
+3. Check Naver Smart Store-specific SEO elements
+4. Verify naver-site-verification meta tag
+5. Check Korean content ratio for Naver visibility
+
+## Output Format
+
+```markdown
+## E-Commerce SEO Audit: [domain]
+
+### Score: [0-100]/100
+
+### Product Page Issues
+- **Critical**: [count] issues
+- **High**: [count] issues
+- **Medium**: [count] issues
+- **Low**: [count] issues
+
+#### Top Issues
+1. [severity] [issue_type] - [message]
+   Recommendation: [fix]
+
+### Category Structure
+- Categories found: [count]
+- Max depth: [number]
+- Breadcrumbs present: [count]
+- Faceted navigation issues: [count]
+
+### Schema Validation
+- Pages with schema: [count]/[total]
+- Valid schemas: [count]
+- Rich result eligible: [count]
+- Common errors: [list]
+
+### Korean Marketplaces
+- Naver Smart Store: [Found/Not Found]
+- Coupang: [Found/Not Found]
+- Gmarket: [Found/Not Found]
+- 11번가: [Found/Not Found]
+
+### Recommendations
+1. [Priority fixes ordered by impact]
+```
+
+## Common Issues
+
+| Issue | Impact | Fix |
+|-------|--------|-----|
+| Missing Product schema | High | Add JSON-LD Product with offers |
+| No canonical on product variants | High | Add self-referencing canonical |
+| Images without alt text | High | Add product name to alt text |
+| Category depth > 4 levels | Medium | Flatten taxonomy |
+| Missing breadcrumbs | Medium | Add BreadcrumbList schema and visible nav |
+| Faceted nav creating duplicates | High | Use canonical or noindex on filtered pages |
+| Missing Naver verification | Medium | Add naver-site-verification meta tag |
+| Price not in KRW for Korean market | Medium | Add KRW pricing to schema |
+
+## Limitations
+
+- Cannot access logged-in areas (member-only products)
+- Marketplace search results may vary by region/IP
+- Large catalogs require sampling (default 50 pages)
+- Cannot validate JavaScript-rendered product content without headless browser
+
+## Notion Output (Required)
+
+All audit reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category (E-Commerce SEO), Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: ECOM-YYYYMMDD-NNN
diff --git a/custom-skills/24-seo-ecommerce/desktop/skill.yaml b/custom-skills/24-seo-ecommerce/desktop/skill.yaml
new file mode 100644
index 0000000..3b29eeb
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-ecommerce
+description: |
+  E-commerce SEO audit and optimization. Triggers: product SEO, e-commerce audit, product schema, category SEO, Smart Store, marketplace SEO.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/24-seo-ecommerce/desktop/tools/ahrefs.md b/custom-skills/24-seo-ecommerce/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..5d918e0
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/desktop/tools/ahrefs.md
@@ -0,0 +1,15 @@
+# Ahrefs
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/24-seo-ecommerce/desktop/tools/notion.md b/custom-skills/24-seo-ecommerce/desktop/tools/notion.md
new file mode 100644
index 0000000..153e086
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/desktop/tools/notion.md
@@ -0,0 +1,15 @@
+# Notion
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/24-seo-ecommerce/desktop/tools/websearch.md b/custom-skills/24-seo-ecommerce/desktop/tools/websearch.md
new file mode 100644
index 0000000..36ef238
--- /dev/null
+++ b/custom-skills/24-seo-ecommerce/desktop/tools/websearch.md
@@ -0,0 +1,15 @@
+# WebSearch
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/25-seo-kpi-framework/code/CLAUDE.md b/custom-skills/25-seo-kpi-framework/code/CLAUDE.md
new file mode 100644
index 0000000..896a9a0
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/code/CLAUDE.md
@@ -0,0 +1,148 @@
+# CLAUDE.md
+
+## Overview
+
+SEO KPI and performance framework for unified metrics aggregation across all SEO dimensions. Establishes baselines, sets targets (30/60/90-day), generates executive summaries with health scores, provides tactical breakdowns, estimates ROI using Ahrefs traffic cost, and supports period-over-period comparison (MoM, QoQ).
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Aggregate KPIs
+python scripts/kpi_aggregator.py --url https://example.com --json
+
+# Generate performance report
+python scripts/performance_reporter.py --url https://example.com --period monthly --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `kpi_aggregator.py` | Aggregate KPIs across all SEO dimensions | Unified KPI dashboard, health score, baselines |
+| `performance_reporter.py` | Generate period-over-period performance reports | Trend analysis, executive summary, tactical breakdown |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## KPI Aggregator
+
+```bash
+# Full KPI aggregation
+python scripts/kpi_aggregator.py --url https://example.com --json
+
+# Set baselines
+python scripts/kpi_aggregator.py --url https://example.com --set-baseline --json
+
+# Compare against baseline
+python scripts/kpi_aggregator.py --url https://example.com --baseline baseline.json --json
+
+# With ROI estimation
+python scripts/kpi_aggregator.py --url https://example.com --roi --json
+```
+
+**Capabilities**:
+- Unified KPI taxonomy across 7 dimensions:
+  - Traffic KPIs (organic sessions, organic traffic value, traffic trend)
+  - Ranking KPIs (visibility score, avg position, top10 keywords count)
+  - Engagement KPIs (bounce rate, pages/session, avg session duration)
+  - Technical KPIs (crawl errors, page speed score, mobile usability)
+  - Content KPIs (indexed pages, content freshness score, thin content ratio)
+  - Link KPIs (domain rating, referring domains, link velocity)
+  - Local KPIs (GBP visibility, review score, citation accuracy)
+- Multi-source data aggregation from Ahrefs and other skill outputs
+- Baseline establishment and target setting (30/60/90 day)
+- Overall health score (0-100) with weighted dimensions
+- ROI estimation using Ahrefs organic traffic cost
+
+## Performance Reporter
+
+```bash
+# Monthly report
+python scripts/performance_reporter.py --url https://example.com --period monthly --json
+
+# Quarterly report
+python scripts/performance_reporter.py --url https://example.com --period quarterly --json
+
+# Custom date range
+python scripts/performance_reporter.py --url https://example.com --from 2025-01-01 --to 2025-03-31 --json
+
+# Executive summary only
+python scripts/performance_reporter.py --url https://example.com --period monthly --executive --json
+```
+
+**Capabilities**:
+- Executive summary generation (health score, trend arrows, key wins/concerns)
+- Period-over-period comparison (MoM, QoQ, YoY)
+- Trend direction indicators (up improving, down declining, stable)
+- Top wins and concerns identification
+- Tactical breakdown with actionable next steps
+- Target vs actual comparison with progress %
+- Traffic value change (ROI proxy)
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `site-explorer-metrics` | Current organic metrics |
+| `site-explorer-metrics-history` | Historical metrics trends |
+| `site-explorer-metrics-by-country` | Country-level breakdown |
+| `site-explorer-domain-rating-history` | DR trend over time |
+| `site-explorer-total-search-volume-history` | Total keyword volume trend |
+
+## Output Format
+
+```json
+{
+  "url": "https://example.com",
+  "health_score": 72,
+  "health_trend": "improving",
+  "kpis": {
+    "traffic": {"organic_traffic": 15000, "traffic_value_usd": 45000, "trend": "up"},
+    "rankings": {"visibility_score": 68, "avg_position": 18.5, "top10_count": 48},
+    "links": {"domain_rating": 45, "referring_domains": 850, "velocity": "+15/month"},
+    "technical": {"health_score": 85, "crawl_errors": 12},
+    "content": {"indexed_pages": 320, "freshness_score": 70},
+    "engagement": {"bounce_rate": 45, "pages_per_session": 2.8},
+    "local": {"gbp_visibility": 80, "review_score": 4.5}
+  },
+  "targets": {
+    "30_day": {},
+    "60_day": {},
+    "90_day": {}
+  },
+  "executive_summary": {
+    "top_wins": [],
+    "top_concerns": [],
+    "recommendations": []
+  },
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Audited website URL |
+| Category | Select | SEO KPI & Performance |
+| Priority | Select | Based on health score trend |
+| Found Date | Date | Report date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: KPI-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., KPI, ROI, Domain Rating, Visibility Score)
+- URLs and code remain unchanged
diff --git a/custom-skills/25-seo-kpi-framework/code/scripts/base_client.py b/custom-skills/25-seo-kpi-framework/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/25-seo-kpi-framework/code/scripts/kpi_aggregator.py b/custom-skills/25-seo-kpi-framework/code/scripts/kpi_aggregator.py
new file mode 100644
index 0000000..1eb21cc
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/code/scripts/kpi_aggregator.py
@@ -0,0 +1,758 @@
+"""
+KPI Aggregator - Unified SEO KPI aggregation across all dimensions
+==================================================================
+Purpose: Aggregate KPIs from Ahrefs and other sources into a unified
+         dashboard with health scores, baselines, targets, and ROI.
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from pathlib import Path
+from typing import Any
+
+import aiohttp
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class KpiMetric:
+    """Single KPI metric with trend and target information."""
+    name: str
+    value: float
+    previous_value: float | None = None
+    change_pct: float | None = None
+    trend: str = "stable"  # up, down, stable
+    target_30d: float | None = None
+    target_60d: float | None = None
+    target_90d: float | None = None
+
+    def compute_trend(self) -> None:
+        """Compute trend direction and change percentage."""
+        if self.previous_value is not None and self.previous_value != 0:
+            self.change_pct = round(
+                ((self.value - self.previous_value) / abs(self.previous_value)) * 100, 2
+            )
+            if self.change_pct > 2.0:
+                self.trend = "up"
+            elif self.change_pct < -2.0:
+                self.trend = "down"
+            else:
+                self.trend = "stable"
+
+
+@dataclass
+class KpiDimension:
+    """A dimension grouping multiple KPI metrics."""
+    name: str
+    metrics: list[KpiMetric] = field(default_factory=list)
+    weight: float = 0.0
+    score: float = 0.0
+
+    def compute_score(self) -> float:
+        """Compute dimension score (0-100) based on metrics health."""
+        if not self.metrics:
+            self.score = 0.0
+            return self.score
+        metric_scores = []
+        for m in self.metrics:
+            if m.trend == "up":
+                metric_scores.append(80.0)
+            elif m.trend == "stable":
+                metric_scores.append(60.0)
+            else:
+                metric_scores.append(35.0)
+            # Boost score if value is positive and non-zero
+            if m.value and m.value > 0:
+                metric_scores[-1] = min(100.0, metric_scores[-1] + 10.0)
+        self.score = round(sum(metric_scores) / len(metric_scores), 1)
+        return self.score
+
+
+@dataclass
+class HealthScore:
+    """Overall SEO health score."""
+    overall: float = 0.0
+    dimensions: dict[str, float] = field(default_factory=dict)
+    trend: str = "stable"
+
+
+@dataclass
+class RoiEstimate:
+    """ROI estimation from Ahrefs traffic cost."""
+    traffic_value_usd: float = 0.0
+    traffic_value_change: float = 0.0
+    estimated_monthly_value: float = 0.0
+
+
+@dataclass
+class KpiResult:
+    """Complete KPI aggregation result."""
+    url: str = ""
+    health_score: float = 0.0
+    health_trend: str = "stable"
+    kpis: dict[str, Any] = field(default_factory=dict)
+    targets: dict[str, Any] = field(default_factory=dict)
+    roi: RoiEstimate | None = None
+    baseline_comparison: dict[str, Any] | None = None
+    executive_summary: dict[str, Any] = field(default_factory=dict)
+    timestamp: str = ""
+    errors: list[str] = field(default_factory=list)
+
+
+# ---------------------------------------------------------------------------
+# Dimension weights
+# ---------------------------------------------------------------------------
+
+DIMENSION_WEIGHTS = {
+    "traffic": 0.25,
+    "rankings": 0.20,
+    "technical": 0.20,
+    "content": 0.15,
+    "links": 0.15,
+    "local": 0.05,
+}
+
+
+# ---------------------------------------------------------------------------
+# KPI Aggregator
+# ---------------------------------------------------------------------------
+
+class KpiAggregator(BaseAsyncClient):
+    """Aggregate SEO KPIs across all dimensions from Ahrefs data."""
+
+    AHREFS_BASE = "https://api.ahrefs.com/v3"
+
+    def __init__(self, api_token: str | None = None):
+        super().__init__(max_concurrent=3, requests_per_second=2.0)
+        self.api_token = api_token or config.get_required("AHREFS_API_TOKEN")
+        self.headers = {
+            "Authorization": f"Bearer {self.api_token}",
+            "Accept": "application/json",
+        }
+
+    # ----- Ahrefs API helpers -----
+
+    async def _ahrefs_get(
+        self, session: aiohttp.ClientSession, endpoint: str, params: dict
+    ) -> dict:
+        """Make an authenticated GET request to Ahrefs API."""
+        url = f"{self.AHREFS_BASE}/{endpoint}"
+        async with session.get(url, headers=self.headers, params=params) as resp:
+            if resp.status != 200:
+                text = await resp.text()
+                self.logger.warning(f"Ahrefs {endpoint} returned {resp.status}: {text}")
+                return {"error": f"HTTP {resp.status}", "detail": text}
+            return await resp.json()
+
+    # ----- Dimension collectors -----
+
+    async def get_traffic_kpis(
+        self, session: aiohttp.ClientSession, url: str
+    ) -> KpiDimension:
+        """Collect traffic KPIs via site-explorer-metrics."""
+        dim = KpiDimension(name="traffic", weight=DIMENSION_WEIGHTS["traffic"])
+        try:
+            data = await self._ahrefs_get(
+                session,
+                "site-explorer/metrics",
+                {"target": url, "mode": "domain"},
+            )
+            if "error" not in data:
+                metrics = data.get("metrics", data)
+                organic = metrics.get("organic", {})
+                organic_traffic = organic.get("traffic", 0)
+                traffic_value_raw = organic.get("cost", 0)
+                traffic_value_usd = traffic_value_raw / 100.0 if traffic_value_raw else 0.0
+                dim.metrics.append(
+                    KpiMetric(name="organic_traffic", value=float(organic_traffic))
+                )
+                dim.metrics.append(
+                    KpiMetric(name="traffic_value_usd", value=round(traffic_value_usd, 2))
+                )
+            else:
+                dim.metrics.append(KpiMetric(name="organic_traffic", value=0.0))
+                dim.metrics.append(KpiMetric(name="traffic_value_usd", value=0.0))
+        except Exception as exc:
+            self.logger.error(f"Traffic KPI error: {exc}")
+            dim.metrics.append(KpiMetric(name="organic_traffic", value=0.0))
+        dim.compute_score()
+        return dim
+
+    async def get_ranking_kpis(
+        self, session: aiohttp.ClientSession, url: str
+    ) -> KpiDimension:
+        """Collect ranking KPIs via site-explorer-metrics."""
+        dim = KpiDimension(name="rankings", weight=DIMENSION_WEIGHTS["rankings"])
+        try:
+            data = await self._ahrefs_get(
+                session,
+                "site-explorer/metrics",
+                {"target": url, "mode": "domain"},
+            )
+            if "error" not in data:
+                metrics = data.get("metrics", data)
+                organic = metrics.get("organic", {})
+                keywords_total = organic.get("keywords", 0)
+                # Estimate top10 as ~20% of total keywords
+                top10_estimate = int(keywords_total * 0.20)
+                # Visibility score heuristic: based on traffic relative to keywords
+                traffic = organic.get("traffic", 0)
+                visibility = min(100.0, (traffic / max(keywords_total, 1)) * 10)
+                dim.metrics.append(
+                    KpiMetric(name="visibility_score", value=round(visibility, 1))
+                )
+                dim.metrics.append(
+                    KpiMetric(name="top10_keywords", value=float(top10_estimate))
+                )
+                dim.metrics.append(
+                    KpiMetric(name="total_keywords", value=float(keywords_total))
+                )
+            else:
+                dim.metrics.append(KpiMetric(name="visibility_score", value=0.0))
+                dim.metrics.append(KpiMetric(name="top10_keywords", value=0.0))
+        except Exception as exc:
+            self.logger.error(f"Ranking KPI error: {exc}")
+            dim.metrics.append(KpiMetric(name="visibility_score", value=0.0))
+        dim.compute_score()
+        return dim
+
+    async def get_link_kpis(
+        self, session: aiohttp.ClientSession, url: str
+    ) -> KpiDimension:
+        """Collect link KPIs via domain-rating and metrics."""
+        dim = KpiDimension(name="links", weight=DIMENSION_WEIGHTS["links"])
+        try:
+            # Domain rating
+            dr_data = await self._ahrefs_get(
+                session,
+                "site-explorer/domain-rating",
+                {"target": url},
+            )
+            domain_rating = 0.0
+            if "error" not in dr_data:
+                domain_rating = float(
+                    dr_data.get("domain_rating", dr_data.get("domainRating", 0))
+                )
+            dim.metrics.append(
+                KpiMetric(name="domain_rating", value=round(domain_rating, 1))
+            )
+
+            # Referring domains from metrics
+            metrics_data = await self._ahrefs_get(
+                session,
+                "site-explorer/metrics",
+                {"target": url, "mode": "domain"},
+            )
+            ref_domains = 0
+            if "error" not in metrics_data:
+                metrics = metrics_data.get("metrics", metrics_data)
+                ref_domains = metrics.get("refdomains", 0)
+            dim.metrics.append(
+                KpiMetric(name="referring_domains", value=float(ref_domains))
+            )
+        except Exception as exc:
+            self.logger.error(f"Link KPI error: {exc}")
+            dim.metrics.append(KpiMetric(name="domain_rating", value=0.0))
+            dim.metrics.append(KpiMetric(name="referring_domains", value=0.0))
+        dim.compute_score()
+        return dim
+
+    async def get_technical_kpis(
+        self, session: aiohttp.ClientSession, url: str
+    ) -> KpiDimension:
+        """Collect technical KPIs (estimated from available data)."""
+        dim = KpiDimension(name="technical", weight=DIMENSION_WEIGHTS["technical"])
+        try:
+            data = await self._ahrefs_get(
+                session,
+                "site-explorer/metrics",
+                {"target": url, "mode": "domain"},
+            )
+            if "error" not in data:
+                metrics = data.get("metrics", data)
+                organic = metrics.get("organic", {})
+                pages_crawled = metrics.get("pages", organic.get("pages", 0))
+                # Heuristic: technical health score from available data
+                has_traffic = organic.get("traffic", 0) > 0
+                has_pages = pages_crawled > 0
+                tech_score = 50.0
+                if has_traffic:
+                    tech_score += 25.0
+                if has_pages:
+                    tech_score += 25.0
+                dim.metrics.append(
+                    KpiMetric(name="technical_health_score", value=round(tech_score, 1))
+                )
+                dim.metrics.append(
+                    KpiMetric(name="pages_crawled", value=float(pages_crawled))
+                )
+            else:
+                dim.metrics.append(KpiMetric(name="technical_health_score", value=50.0))
+                dim.metrics.append(KpiMetric(name="pages_crawled", value=0.0))
+        except Exception as exc:
+            self.logger.error(f"Technical KPI error: {exc}")
+            dim.metrics.append(KpiMetric(name="technical_health_score", value=50.0))
+        dim.compute_score()
+        return dim
+
+    async def get_content_kpis(
+        self, session: aiohttp.ClientSession, url: str
+    ) -> KpiDimension:
+        """Collect content KPIs from available metrics."""
+        dim = KpiDimension(name="content", weight=DIMENSION_WEIGHTS["content"])
+        try:
+            data = await self._ahrefs_get(
+                session,
+                "site-explorer/metrics",
+                {"target": url, "mode": "domain"},
+            )
+            if "error" not in data:
+                metrics = data.get("metrics", data)
+                organic = metrics.get("organic", {})
+                pages = metrics.get("pages", organic.get("pages", 0))
+                keywords = organic.get("keywords", 0)
+                # Content freshness heuristic
+                freshness = min(100.0, (keywords / max(pages, 1)) * 5) if pages else 0.0
+                dim.metrics.append(
+                    KpiMetric(name="indexed_pages", value=float(pages))
+                )
+                dim.metrics.append(
+                    KpiMetric(name="content_freshness_score", value=round(freshness, 1))
+                )
+                dim.metrics.append(
+                    KpiMetric(name="keywords_per_page", value=round(keywords / max(pages, 1), 2))
+                )
+            else:
+                dim.metrics.append(KpiMetric(name="indexed_pages", value=0.0))
+                dim.metrics.append(KpiMetric(name="content_freshness_score", value=0.0))
+        except Exception as exc:
+            self.logger.error(f"Content KPI error: {exc}")
+            dim.metrics.append(KpiMetric(name="indexed_pages", value=0.0))
+        dim.compute_score()
+        return dim
+
+    async def get_local_kpis(self, url: str) -> KpiDimension:
+        """Placeholder for local KPIs (requires external data)."""
+        dim = KpiDimension(name="local", weight=DIMENSION_WEIGHTS["local"])
+        dim.metrics.append(KpiMetric(name="gbp_visibility", value=0.0))
+        dim.metrics.append(KpiMetric(name="review_score", value=0.0))
+        dim.metrics.append(KpiMetric(name="citation_accuracy", value=0.0))
+        dim.compute_score()
+        return dim
+
+    # ----- Health score -----
+
+    def calculate_health_score(self, dimensions: list[KpiDimension]) -> HealthScore:
+        """Calculate weighted health score across all dimensions."""
+        health = HealthScore()
+        total_weight = 0.0
+        weighted_sum = 0.0
+
+        for dim in dimensions:
+            dim.compute_score()
+            health.dimensions[dim.name] = dim.score
+            weighted_sum += dim.score * dim.weight
+            total_weight += dim.weight
+
+        if total_weight > 0:
+            health.overall = round(weighted_sum / total_weight, 1)
+        else:
+            health.overall = 0.0
+
+        # Determine trend from dimension trends
+        up_count = sum(
+            1 for d in dimensions
+            for m in d.metrics if m.trend == "up"
+        )
+        down_count = sum(
+            1 for d in dimensions
+            for m in d.metrics if m.trend == "down"
+        )
+        if up_count > down_count:
+            health.trend = "improving"
+        elif down_count > up_count:
+            health.trend = "declining"
+        else:
+            health.trend = "stable"
+
+        return health
+
+    # ----- Targets -----
+
+    def set_targets(self, dimensions: list[KpiDimension]) -> dict[str, Any]:
+        """Calculate 30/60/90 day targets (5%/10%/20% improvement)."""
+        targets = {"30_day": {}, "60_day": {}, "90_day": {}}
+        growth_rates = {"30_day": 0.05, "60_day": 0.10, "90_day": 0.20}
+
+        for dim in dimensions:
+            for metric in dim.metrics:
+                if metric.value and metric.value > 0:
+                    for period, rate in growth_rates.items():
+                        key = f"{dim.name}.{metric.name}"
+                        # For metrics where lower is better (e.g. bounce rate),
+                        # improvement means decrease
+                        if metric.name in ("bounce_rate", "crawl_errors", "thin_content_ratio"):
+                            target_val = metric.value * (1 - rate)
+                        else:
+                            target_val = metric.value * (1 + rate)
+                        targets[period][key] = round(target_val, 2)
+                    metric.target_30d = targets["30_day"].get(f"{dim.name}.{metric.name}")
+                    metric.target_60d = targets["60_day"].get(f"{dim.name}.{metric.name}")
+                    metric.target_90d = targets["90_day"].get(f"{dim.name}.{metric.name}")
+        return targets
+
+    # ----- ROI estimation -----
+
+    def estimate_roi(self, traffic_dim: KpiDimension) -> RoiEstimate:
+        """Estimate ROI from Ahrefs traffic cost data."""
+        roi = RoiEstimate()
+        for metric in traffic_dim.metrics:
+            if metric.name == "traffic_value_usd":
+                roi.traffic_value_usd = metric.value
+                roi.estimated_monthly_value = metric.value
+                if metric.previous_value is not None:
+                    roi.traffic_value_change = round(
+                        metric.value - metric.previous_value, 2
+                    )
+        return roi
+
+    # ----- Baseline comparison -----
+
+    def compare_baseline(
+        self, current: list[KpiDimension], baseline: dict[str, Any]
+    ) -> dict[str, Any]:
+        """Compare current KPIs against a stored baseline."""
+        comparison = {}
+        baseline_kpis = baseline.get("kpis", {})
+
+        for dim in current:
+            dim_baseline = baseline_kpis.get(dim.name, {})
+            dim_comparison = {}
+            for metric in dim.metrics:
+                baseline_val = None
+                if isinstance(dim_baseline, dict):
+                    baseline_val = dim_baseline.get(metric.name)
+                if baseline_val is not None:
+                    metric.previous_value = float(baseline_val)
+                    metric.compute_trend()
+                    dim_comparison[metric.name] = {
+                        "current": metric.value,
+                        "baseline": baseline_val,
+                        "change_pct": metric.change_pct,
+                        "trend": metric.trend,
+                    }
+                else:
+                    dim_comparison[metric.name] = {
+                        "current": metric.value,
+                        "baseline": None,
+                        "change_pct": None,
+                        "trend": "no_baseline",
+                    }
+            comparison[dim.name] = dim_comparison
+        return comparison
+
+    # ----- Executive summary -----
+
+    def generate_executive_summary(
+        self, dimensions: list[KpiDimension], health: HealthScore
+    ) -> dict[str, Any]:
+        """Generate executive summary with wins, concerns, recommendations."""
+        wins = []
+        concerns = []
+        recommendations = []
+
+        for dim in dimensions:
+            for metric in dim.metrics:
+                if metric.trend == "up" and metric.change_pct and metric.change_pct > 5:
+                    wins.append(
+                        f"{dim.name}/{metric.name}: +{metric.change_pct}% improvement"
+                    )
+                elif metric.trend == "down" and metric.change_pct and metric.change_pct < -5:
+                    concerns.append(
+                        f"{dim.name}/{metric.name}: {metric.change_pct}% decline"
+                    )
+
+        # Generate recommendations based on dimension scores
+        for dim in dimensions:
+            if dim.score < 50:
+                recommendations.append(
+                    f"Priority: Improve {dim.name} dimension (score: {dim.score}/100)"
+                )
+            elif dim.score < 70:
+                recommendations.append(
+                    f"Monitor: {dim.name} dimension needs attention (score: {dim.score}/100)"
+                )
+
+        if not wins:
+            wins.append("No significant improvements detected in this period")
+        if not concerns:
+            concerns.append("No significant declines detected in this period")
+        if not recommendations:
+            recommendations.append("All dimensions performing well - maintain current strategy")
+
+        return {
+            "health_score": health.overall,
+            "health_trend": health.trend,
+            "top_wins": wins[:5],
+            "top_concerns": concerns[:5],
+            "recommendations": recommendations[:5],
+        }
+
+    # ----- Main orchestration -----
+
+    async def aggregate(
+        self,
+        url: str,
+        include_roi: bool = False,
+        baseline_path: str | None = None,
+        set_baseline: bool = False,
+    ) -> KpiResult:
+        """Orchestrate full KPI aggregation across all dimensions."""
+        result = KpiResult(url=url, timestamp=datetime.now().isoformat())
+        dimensions: list[KpiDimension] = []
+
+        async with aiohttp.ClientSession() as session:
+            # Collect all dimensions concurrently
+            tasks = [
+                self.get_traffic_kpis(session, url),
+                self.get_ranking_kpis(session, url),
+                self.get_link_kpis(session, url),
+                self.get_technical_kpis(session, url),
+                self.get_content_kpis(session, url),
+            ]
+            gathered = await asyncio.gather(*tasks, return_exceptions=True)
+
+            for item in gathered:
+                if isinstance(item, Exception):
+                    result.errors.append(str(item))
+                    self.logger.error(f"Dimension error: {item}")
+                else:
+                    dimensions.append(item)
+
+            # Local KPIs (no API call needed)
+            local_dim = await self.get_local_kpis(url)
+            dimensions.append(local_dim)
+
+        # Load baseline if provided
+        if baseline_path:
+            try:
+                baseline_data = json.loads(Path(baseline_path).read_text())
+                result.baseline_comparison = self.compare_baseline(dimensions, baseline_data)
+            except Exception as exc:
+                result.errors.append(f"Baseline load error: {exc}")
+
+        # Calculate health score
+        health = self.calculate_health_score(dimensions)
+        result.health_score = health.overall
+        result.health_trend = health.trend
+
+        # Build KPI dictionary
+        for dim in dimensions:
+            result.kpis[dim.name] = {
+                "score": dim.score,
+                "weight": dim.weight,
+                "metrics": {m.name: asdict(m) for m in dim.metrics},
+            }
+
+        # Set targets
+        targets = self.set_targets(dimensions)
+        result.targets = targets
+
+        # ROI estimation
+        if include_roi:
+            traffic_dim = next((d for d in dimensions if d.name == "traffic"), None)
+            if traffic_dim:
+                roi = self.estimate_roi(traffic_dim)
+                result.roi = roi
+
+        # Executive summary
+        result.executive_summary = self.generate_executive_summary(dimensions, health)
+
+        # Save baseline if requested
+        if set_baseline:
+            baseline_out = {
+                "url": url,
+                "timestamp": result.timestamp,
+                "kpis": {},
+            }
+            for dim in dimensions:
+                baseline_out["kpis"][dim.name] = {
+                    m.name: m.value for m in dim.metrics
+                }
+            baseline_file = f"baseline_{url.replace('https://', '').replace('/', '_')}.json"
+            Path(baseline_file).write_text(json.dumps(baseline_out, indent=2))
+            self.logger.info(f"Baseline saved to {baseline_file}")
+
+        return result
+
+
+# ---------------------------------------------------------------------------
+# Output formatting
+# ---------------------------------------------------------------------------
+
+def format_text_report(result: KpiResult) -> str:
+    """Format KPI result as human-readable text report."""
+    lines = []
+    lines.append("=" * 60)
+    lines.append(f"SEO KPI Dashboard: {result.url}")
+    lines.append(f"Timestamp: {result.timestamp}")
+    lines.append("=" * 60)
+    lines.append("")
+
+    # Health score
+    lines.append(f"Overall Health Score: {result.health_score}/100  ({result.health_trend})")
+    lines.append("-" * 40)
+
+    # Dimensions
+    for dim_name, dim_data in result.kpis.items():
+        lines.append(f"\n[{dim_name.upper()}] Score: {dim_data['score']}/100 (weight: {dim_data['weight']})")
+        metrics = dim_data.get("metrics", {})
+        for m_name, m_data in metrics.items():
+            trend_arrow = {"up": "^", "down": "v", "stable": "=", "no_baseline": "?"}.get(
+                m_data.get("trend", "stable"), "="
+            )
+            val = m_data.get("value", 0)
+            change = m_data.get("change_pct")
+            change_str = f" ({change:+.1f}%)" if change is not None else ""
+            lines.append(f"  {trend_arrow} {m_name}: {val}{change_str}")
+
+    # Targets
+    if result.targets:
+        lines.append("\n" + "-" * 40)
+        lines.append("TARGETS")
+        for period, targets in result.targets.items():
+            if targets:
+                lines.append(f"\n  {period}:")
+                for key, val in list(targets.items())[:10]:
+                    lines.append(f"    {key}: {val}")
+
+    # ROI
+    if result.roi:
+        lines.append("\n" + "-" * 40)
+        lines.append("ROI ESTIMATE")
+        lines.append(f"  Traffic Value (USD): ${result.roi.traffic_value_usd:,.2f}")
+        lines.append(f"  Monthly Value: ${result.roi.estimated_monthly_value:,.2f}")
+        lines.append(f"  Value Change: ${result.roi.traffic_value_change:,.2f}")
+
+    # Executive summary
+    if result.executive_summary:
+        lines.append("\n" + "-" * 40)
+        lines.append("EXECUTIVE SUMMARY")
+        lines.append(f"  Health: {result.executive_summary.get('health_score', 0)}/100")
+        lines.append(f"  Trend: {result.executive_summary.get('health_trend', 'stable')}")
+        lines.append("\n  Top Wins:")
+        for win in result.executive_summary.get("top_wins", []):
+            lines.append(f"    + {win}")
+        lines.append("\n  Top Concerns:")
+        for concern in result.executive_summary.get("top_concerns", []):
+            lines.append(f"    - {concern}")
+        lines.append("\n  Recommendations:")
+        for rec in result.executive_summary.get("recommendations", []):
+            lines.append(f"    > {rec}")
+
+    # Errors
+    if result.errors:
+        lines.append("\n" + "-" * 40)
+        lines.append("ERRORS:")
+        for err in result.errors:
+            lines.append(f"  ! {err}")
+
+    lines.append("\n" + "=" * 60)
+    return "\n".join(lines)
+
+
+def serialize_result(result: KpiResult) -> dict:
+    """Serialize KpiResult to JSON-safe dictionary."""
+    data = {
+        "url": result.url,
+        "health_score": result.health_score,
+        "health_trend": result.health_trend,
+        "kpis": result.kpis,
+        "targets": result.targets,
+        "executive_summary": result.executive_summary,
+        "timestamp": result.timestamp,
+        "errors": result.errors,
+    }
+    if result.roi:
+        data["roi"] = asdict(result.roi)
+    if result.baseline_comparison:
+        data["baseline_comparison"] = result.baseline_comparison
+    return data
+
+
+# ---------------------------------------------------------------------------
+# CLI entry point
+# ---------------------------------------------------------------------------
+
+def parse_args() -> argparse.Namespace:
+    """Parse command-line arguments."""
+    parser = argparse.ArgumentParser(
+        description="SEO KPI Aggregator - Unified metrics dashboard"
+    )
+    parser.add_argument(
+        "--url", required=True, help="Target URL or domain to analyze"
+    )
+    parser.add_argument(
+        "--set-baseline", action="store_true",
+        help="Save current KPIs as baseline file"
+    )
+    parser.add_argument(
+        "--baseline", type=str, default=None,
+        help="Path to baseline JSON file for comparison"
+    )
+    parser.add_argument(
+        "--roi", action="store_true",
+        help="Include ROI estimation from traffic cost"
+    )
+    parser.add_argument(
+        "--json", action="store_true",
+        help="Output results as JSON"
+    )
+    parser.add_argument(
+        "--output", type=str, default=None,
+        help="Save output to file path"
+    )
+    return parser.parse_args()
+
+
+async def main() -> None:
+    """Main entry point."""
+    args = parse_args()
+
+    aggregator = KpiAggregator()
+    result = await aggregator.aggregate(
+        url=args.url,
+        include_roi=args.roi,
+        baseline_path=args.baseline,
+        set_baseline=args.set_baseline,
+    )
+
+    if args.json:
+        output = json.dumps(serialize_result(result), indent=2, ensure_ascii=False)
+    else:
+        output = format_text_report(result)
+
+    if args.output:
+        Path(args.output).write_text(output, encoding="utf-8")
+        logger.info(f"Output saved to {args.output}")
+    else:
+        print(output)
+
+    aggregator.print_stats()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/25-seo-kpi-framework/code/scripts/performance_reporter.py b/custom-skills/25-seo-kpi-framework/code/scripts/performance_reporter.py
new file mode 100644
index 0000000..27f42d9
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/code/scripts/performance_reporter.py
@@ -0,0 +1,801 @@
+"""
+Performance Reporter - Period-over-period SEO performance reports
+================================================================
+Purpose: Generate executive summaries, trend analysis, tactical breakdowns,
+         and target-vs-actual comparison from Ahrefs historical data.
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime, timedelta
+from pathlib import Path
+from typing import Any
+
+import aiohttp
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class TrendData:
+    """Single trend data point for a metric."""
+    period: str
+    value: float
+    change_pct: float | None = None
+    direction: str = "stable"  # up, down, stable
+
+
+@dataclass
+class WinConcern:
+    """A notable win or concern from performance analysis."""
+    category: str
+    description: str
+    impact: str = "medium"  # high, medium, low
+    action: str = ""
+
+
+@dataclass
+class TargetProgress:
+    """Target vs actual progress tracking."""
+    kpi_name: str
+    target: float
+    actual: float
+    progress_pct: float = 0.0
+
+    def compute_progress(self) -> None:
+        """Compute progress percentage toward target."""
+        if self.target and self.target != 0:
+            self.progress_pct = round((self.actual / self.target) * 100, 1)
+        else:
+            self.progress_pct = 0.0
+
+
+@dataclass
+class PerformanceReport:
+    """Complete performance report."""
+    url: str = ""
+    period: str = "monthly"
+    date_from: str = ""
+    date_to: str = ""
+    health_score: float = 0.0
+    health_trend: str = "stable"
+    trends: dict[str, list[TrendData]] = field(default_factory=dict)
+    wins: list[WinConcern] = field(default_factory=list)
+    concerns: list[WinConcern] = field(default_factory=list)
+    executive_summary: dict[str, Any] = field(default_factory=dict)
+    tactical_breakdown: dict[str, Any] = field(default_factory=dict)
+    target_progress: list[TargetProgress] = field(default_factory=list)
+    traffic_value_change: float = 0.0
+    timestamp: str = ""
+    errors: list[str] = field(default_factory=list)
+
+
+# ---------------------------------------------------------------------------
+# Period helpers
+# ---------------------------------------------------------------------------
+
+PERIOD_DAYS = {
+    "monthly": 30,
+    "quarterly": 90,
+    "yearly": 365,
+}
+
+
+def get_date_range(
+    period: str, date_from: str | None = None, date_to: str | None = None
+) -> tuple[str, str]:
+    """Compute date range from period or explicit dates."""
+    if date_from and date_to:
+        return date_from, date_to
+    end = datetime.now()
+    days = PERIOD_DAYS.get(period, 30)
+    start = end - timedelta(days=days)
+    return start.strftime("%Y-%m-%d"), end.strftime("%Y-%m-%d")
+
+
+def get_previous_range(
+    date_from: str, date_to: str
+) -> tuple[str, str]:
+    """Compute the previous period of equal length for comparison."""
+    start = datetime.strptime(date_from, "%Y-%m-%d")
+    end = datetime.strptime(date_to, "%Y-%m-%d")
+    delta = end - start
+    prev_end = start - timedelta(days=1)
+    prev_start = prev_end - delta
+    return prev_start.strftime("%Y-%m-%d"), prev_end.strftime("%Y-%m-%d")
+
+
+# ---------------------------------------------------------------------------
+# Performance Reporter
+# ---------------------------------------------------------------------------
+
+class PerformanceReporter(BaseAsyncClient):
+    """Generate period-over-period SEO performance reports from Ahrefs."""
+
+    AHREFS_BASE = "https://api.ahrefs.com/v3"
+
+    def __init__(self, api_token: str | None = None):
+        super().__init__(max_concurrent=3, requests_per_second=2.0)
+        self.api_token = api_token or config.get_required("AHREFS_API_TOKEN")
+        self.headers = {
+            "Authorization": f"Bearer {self.api_token}",
+            "Accept": "application/json",
+        }
+
+    async def _ahrefs_get(
+        self, session: aiohttp.ClientSession, endpoint: str, params: dict
+    ) -> dict:
+        """Make an authenticated GET request to Ahrefs API."""
+        url = f"{self.AHREFS_BASE}/{endpoint}"
+        async with session.get(url, headers=self.headers, params=params) as resp:
+            if resp.status != 200:
+                text = await resp.text()
+                self.logger.warning(f"Ahrefs {endpoint} returned {resp.status}: {text}")
+                return {"error": f"HTTP {resp.status}", "detail": text}
+            return await resp.json()
+
+    # ----- Data collectors -----
+
+    async def get_metrics_history(
+        self,
+        session: aiohttp.ClientSession,
+        url: str,
+        date_from: str,
+        date_to: str,
+    ) -> list[dict]:
+        """Fetch historical metrics via site-explorer-metrics-history."""
+        data = await self._ahrefs_get(
+            session,
+            "site-explorer/metrics-history",
+            {
+                "target": url,
+                "mode": "domain",
+                "date_from": date_from,
+                "date_to": date_to,
+            },
+        )
+        if "error" in data:
+            self.logger.warning(f"Metrics history error: {data}")
+            return []
+        return data.get("metrics", data.get("data", []))
+
+    async def get_dr_history(
+        self,
+        session: aiohttp.ClientSession,
+        url: str,
+        date_from: str,
+        date_to: str,
+    ) -> list[dict]:
+        """Fetch domain rating history."""
+        data = await self._ahrefs_get(
+            session,
+            "site-explorer/domain-rating-history",
+            {
+                "target": url,
+                "date_from": date_from,
+                "date_to": date_to,
+            },
+        )
+        if "error" in data:
+            return []
+        return data.get("domain_rating_history", data.get("data", []))
+
+    async def get_current_metrics(
+        self, session: aiohttp.ClientSession, url: str
+    ) -> dict:
+        """Fetch current snapshot metrics."""
+        data = await self._ahrefs_get(
+            session,
+            "site-explorer/metrics",
+            {"target": url, "mode": "domain"},
+        )
+        if "error" in data:
+            return {}
+        return data.get("metrics", data)
+
+    async def get_volume_history(
+        self,
+        session: aiohttp.ClientSession,
+        url: str,
+        date_from: str,
+        date_to: str,
+    ) -> list[dict]:
+        """Fetch total search volume history."""
+        data = await self._ahrefs_get(
+            session,
+            "site-explorer/total-search-volume-history",
+            {
+                "target": url,
+                "date_from": date_from,
+                "date_to": date_to,
+            },
+        )
+        if "error" in data:
+            return []
+        return data.get("total_search_volume_history", data.get("data", []))
+
+    # ----- Analysis methods -----
+
+    def calculate_period_comparison(
+        self, current_data: list[dict], previous_data: list[dict], metric_key: str
+    ) -> list[TrendData]:
+        """Compare metric values between current and previous period."""
+        trends = []
+
+        def avg_metric(data_list: list[dict], key: str) -> float:
+            vals = []
+            for entry in data_list:
+                val = entry.get(key)
+                if val is None:
+                    organic = entry.get("organic", {})
+                    val = organic.get(key)
+                if val is not None:
+                    vals.append(float(val))
+            return sum(vals) / len(vals) if vals else 0.0
+
+        current_avg = avg_metric(current_data, metric_key)
+        previous_avg = avg_metric(previous_data, metric_key)
+
+        change_pct = None
+        direction = "stable"
+        if previous_avg and previous_avg != 0:
+            change_pct = round(((current_avg - previous_avg) / abs(previous_avg)) * 100, 2)
+            if change_pct > 2.0:
+                direction = "up"
+            elif change_pct < -2.0:
+                direction = "down"
+
+        trends.append(TrendData(
+            period=metric_key,
+            value=round(current_avg, 2),
+            change_pct=change_pct,
+            direction=direction,
+        ))
+        return trends
+
+    def identify_wins(
+        self, current: dict, previous: dict
+    ) -> list[WinConcern]:
+        """Identify significant positive changes between periods."""
+        wins = []
+        metric_labels = {
+            "traffic": "Organic Traffic",
+            "cost": "Traffic Value",
+            "keywords": "Keyword Count",
+            "refdomains": "Referring Domains",
+        }
+
+        for key, label in metric_labels.items():
+            curr_val = self._extract_metric(current, key)
+            prev_val = self._extract_metric(previous, key)
+            if prev_val and prev_val > 0 and curr_val > prev_val:
+                change_pct = ((curr_val - prev_val) / prev_val) * 100
+                if change_pct >= 5.0:
+                    impact = "high" if change_pct >= 20 else ("medium" if change_pct >= 10 else "low")
+                    wins.append(WinConcern(
+                        category=label,
+                        description=f"{label} increased by {change_pct:+.1f}% ({prev_val:,.0f} -> {curr_val:,.0f})",
+                        impact=impact,
+                        action=f"Continue current {label.lower()} strategy",
+                    ))
+        return wins
+
+    def identify_concerns(
+        self, current: dict, previous: dict
+    ) -> list[WinConcern]:
+        """Identify significant negative changes between periods."""
+        concerns = []
+        metric_labels = {
+            "traffic": "Organic Traffic",
+            "cost": "Traffic Value",
+            "keywords": "Keyword Count",
+            "refdomains": "Referring Domains",
+        }
+
+        for key, label in metric_labels.items():
+            curr_val = self._extract_metric(current, key)
+            prev_val = self._extract_metric(previous, key)
+            if prev_val and prev_val > 0 and curr_val < prev_val:
+                change_pct = ((curr_val - prev_val) / prev_val) * 100
+                if change_pct <= -5.0:
+                    impact = "high" if change_pct <= -20 else ("medium" if change_pct <= -10 else "low")
+                    actions = {
+                        "Organic Traffic": "Investigate traffic sources and algorithm updates",
+                        "Traffic Value": "Review keyword targeting and content quality",
+                        "Keyword Count": "Expand content coverage and optimize existing pages",
+                        "Referring Domains": "Strengthen link building outreach campaigns",
+                    }
+                    concerns.append(WinConcern(
+                        category=label,
+                        description=f"{label} decreased by {change_pct:.1f}% ({prev_val:,.0f} -> {curr_val:,.0f})",
+                        impact=impact,
+                        action=actions.get(label, f"Review {label.lower()} strategy"),
+                    ))
+        return concerns
+
+    def _extract_metric(self, data: dict, key: str) -> float:
+        """Extract a metric value from nested Ahrefs response."""
+        if key in data:
+            return float(data[key])
+        organic = data.get("organic", {})
+        if key in organic:
+            return float(organic[key])
+        return 0.0
+
+    def generate_executive_summary(
+        self,
+        wins: list[WinConcern],
+        concerns: list[WinConcern],
+        health_score: float,
+        health_trend: str,
+        traffic_value_change: float,
+    ) -> dict[str, Any]:
+        """Generate high-level executive summary."""
+        summary = {
+            "health_score": health_score,
+            "health_trend": health_trend,
+            "traffic_value_change_usd": round(traffic_value_change, 2),
+            "total_wins": len(wins),
+            "total_concerns": len(concerns),
+            "top_wins": [
+                {"category": w.category, "description": w.description, "impact": w.impact}
+                for w in sorted(wins, key=lambda x: {"high": 0, "medium": 1, "low": 2}.get(x.impact, 3))[:5]
+            ],
+            "top_concerns": [
+                {"category": c.category, "description": c.description, "impact": c.impact}
+                for c in sorted(concerns, key=lambda x: {"high": 0, "medium": 1, "low": 2}.get(x.impact, 3))[:5]
+            ],
+            "overall_assessment": "",
+        }
+
+        if health_score >= 75:
+            summary["overall_assessment"] = "Strong performance - focus on maintaining momentum"
+        elif health_score >= 50:
+            summary["overall_assessment"] = "Moderate performance - targeted improvements needed"
+        else:
+            summary["overall_assessment"] = "Needs attention - prioritize fundamental improvements"
+
+        return summary
+
+    def generate_tactical_breakdown(
+        self, current: dict, wins: list[WinConcern], concerns: list[WinConcern]
+    ) -> dict[str, Any]:
+        """Generate actionable next steps per dimension."""
+        breakdown = {
+            "traffic": {
+                "status": "needs_review",
+                "actions": [],
+            },
+            "rankings": {
+                "status": "needs_review",
+                "actions": [],
+            },
+            "links": {
+                "status": "needs_review",
+                "actions": [],
+            },
+            "content": {
+                "status": "needs_review",
+                "actions": [],
+            },
+            "technical": {
+                "status": "needs_review",
+                "actions": [],
+            },
+        }
+
+        traffic = self._extract_metric(current, "traffic")
+        keywords = self._extract_metric(current, "keywords")
+        refdomains = self._extract_metric(current, "refdomains")
+
+        # Traffic actions
+        if traffic > 0:
+            breakdown["traffic"]["status"] = "active"
+            breakdown["traffic"]["actions"].append("Monitor top landing pages for traffic drops")
+            breakdown["traffic"]["actions"].append("Identify new keyword opportunities in adjacent topics")
+        else:
+            breakdown["traffic"]["actions"].append("Establish organic traffic baseline with content strategy")
+
+        # Rankings actions
+        if keywords > 0:
+            breakdown["rankings"]["status"] = "active"
+            breakdown["rankings"]["actions"].append(
+                f"Optimize pages for {int(keywords)} tracked keywords"
+            )
+            breakdown["rankings"]["actions"].append("Target featured snippets for top-performing queries")
+        else:
+            breakdown["rankings"]["actions"].append("Begin keyword research and content mapping")
+
+        # Links actions
+        if refdomains > 0:
+            breakdown["links"]["status"] = "active"
+            breakdown["links"]["actions"].append("Analyze top referring domains for partnership opportunities")
+            breakdown["links"]["actions"].append("Monitor for lost backlinks and reclaim valuable links")
+        else:
+            breakdown["links"]["actions"].append("Develop link acquisition strategy with digital PR")
+
+        # Content actions
+        breakdown["content"]["actions"].append("Audit content freshness and update older pages")
+        breakdown["content"]["actions"].append("Identify content gaps using competitor analysis")
+
+        # Technical actions
+        breakdown["technical"]["actions"].append("Run technical SEO audit for crawl issues")
+        breakdown["technical"]["actions"].append("Verify Core Web Vitals pass thresholds")
+
+        # Enrich with win/concern context
+        for w in wins:
+            cat_lower = w.category.lower()
+            if "traffic" in cat_lower and breakdown.get("traffic"):
+                breakdown["traffic"]["status"] = "improving"
+            if "keyword" in cat_lower and breakdown.get("rankings"):
+                breakdown["rankings"]["status"] = "improving"
+            if "domain" in cat_lower or "link" in cat_lower:
+                breakdown["links"]["status"] = "improving"
+
+        for c in concerns:
+            cat_lower = c.category.lower()
+            if "traffic" in cat_lower and breakdown.get("traffic"):
+                breakdown["traffic"]["status"] = "declining"
+                breakdown["traffic"]["actions"].insert(0, c.action)
+            if "keyword" in cat_lower and breakdown.get("rankings"):
+                breakdown["rankings"]["status"] = "declining"
+                breakdown["rankings"]["actions"].insert(0, c.action)
+
+        return breakdown
+
+    def compare_targets(
+        self, current: dict, targets: dict
+    ) -> list[TargetProgress]:
+        """Compare current metrics against saved targets."""
+        progress_list = []
+        for key, target_val in targets.items():
+            parts = key.split(".")
+            metric_name = parts[-1] if len(parts) > 1 else key
+            actual = self._extract_metric(current, metric_name)
+            if actual == 0.0 and len(parts) > 1:
+                # Try alternate key resolution
+                actual = current.get(key, 0.0)
+                if isinstance(actual, dict):
+                    actual = 0.0
+            tp = TargetProgress(
+                kpi_name=key,
+                target=float(target_val),
+                actual=float(actual),
+            )
+            tp.compute_progress()
+            progress_list.append(tp)
+        return progress_list
+
+    # ----- Main orchestration -----
+
+    async def report(
+        self,
+        url: str,
+        period: str = "monthly",
+        date_from: str | None = None,
+        date_to: str | None = None,
+        executive_only: bool = False,
+        targets_path: str | None = None,
+    ) -> PerformanceReport:
+        """Orchestrate full performance report generation."""
+        report = PerformanceReport(
+            url=url,
+            period=period,
+            timestamp=datetime.now().isoformat(),
+        )
+
+        # Determine date ranges
+        report.date_from, report.date_to = get_date_range(period, date_from, date_to)
+        prev_from, prev_to = get_previous_range(report.date_from, report.date_to)
+
+        async with aiohttp.ClientSession() as session:
+            # Fetch current and previous period data concurrently
+            tasks = [
+                self.get_metrics_history(session, url, report.date_from, report.date_to),
+                self.get_metrics_history(session, url, prev_from, prev_to),
+                self.get_current_metrics(session, url),
+                self.get_dr_history(session, url, report.date_from, report.date_to),
+                self.get_volume_history(session, url, report.date_from, report.date_to),
+            ]
+            results = await asyncio.gather(*tasks, return_exceptions=True)
+
+            current_history = results[0] if not isinstance(results[0], Exception) else []
+            previous_history = results[1] if not isinstance(results[1], Exception) else []
+            current_snapshot = results[2] if not isinstance(results[2], Exception) else {}
+            dr_history = results[3] if not isinstance(results[3], Exception) else []
+            volume_history = results[4] if not isinstance(results[4], Exception) else []
+
+            for i, r in enumerate(results):
+                if isinstance(r, Exception):
+                    report.errors.append(f"Data fetch error [{i}]: {r}")
+
+        # Calculate trends for key metrics
+        for metric_key in ["traffic", "keywords", "cost", "refdomains"]:
+            if current_history or previous_history:
+                trend = self.calculate_period_comparison(
+                    current_history if isinstance(current_history, list) else [],
+                    previous_history if isinstance(previous_history, list) else [],
+                    metric_key,
+                )
+                report.trends[metric_key] = [asdict(t) for t in trend]
+
+        # Build previous snapshot for comparison
+        previous_snapshot = {}
+        if isinstance(previous_history, list) and previous_history:
+            for entry in previous_history:
+                for key in ("traffic", "cost", "keywords", "refdomains"):
+                    val = entry.get(key)
+                    if val is None:
+                        organic = entry.get("organic", {})
+                        val = organic.get(key)
+                    if val is not None:
+                        if key not in previous_snapshot:
+                            previous_snapshot[key] = []
+                        previous_snapshot[key].append(float(val))
+            # Average the values
+            previous_snapshot = {
+                k: sum(v) / len(v) for k, v in previous_snapshot.items() if v
+            }
+
+        # Identify wins and concerns
+        if isinstance(current_snapshot, dict):
+            report.wins = self.identify_wins(current_snapshot, previous_snapshot)
+            report.concerns = self.identify_concerns(current_snapshot, previous_snapshot)
+        else:
+            report.wins = []
+            report.concerns = []
+
+        # Calculate health score (simple heuristic)
+        traffic = self._extract_metric(current_snapshot, "traffic") if isinstance(current_snapshot, dict) else 0
+        keywords = self._extract_metric(current_snapshot, "keywords") if isinstance(current_snapshot, dict) else 0
+        score_components = []
+        if traffic > 0:
+            score_components.append(min(100, traffic / 100))
+        if keywords > 0:
+            score_components.append(min(100, keywords / 50))
+        if dr_history:
+            latest_dr = dr_history[-1] if isinstance(dr_history, list) else {}
+            dr_val = latest_dr.get("domain_rating", latest_dr.get("domainRating", 0))
+            score_components.append(float(dr_val))
+        report.health_score = round(
+            sum(score_components) / max(len(score_components), 1), 1
+        )
+
+        # Health trend
+        win_count = len(report.wins)
+        concern_count = len(report.concerns)
+        if win_count > concern_count:
+            report.health_trend = "improving"
+        elif concern_count > win_count:
+            report.health_trend = "declining"
+        else:
+            report.health_trend = "stable"
+
+        # Traffic value change
+        curr_cost = self._extract_metric(current_snapshot, "cost") if isinstance(current_snapshot, dict) else 0
+        prev_cost = previous_snapshot.get("cost", 0)
+        report.traffic_value_change = round((curr_cost - prev_cost) / 100.0, 2)
+
+        # Executive summary
+        report.executive_summary = self.generate_executive_summary(
+            report.wins, report.concerns,
+            report.health_score, report.health_trend,
+            report.traffic_value_change,
+        )
+
+        if not executive_only:
+            # Tactical breakdown
+            report.tactical_breakdown = self.generate_tactical_breakdown(
+                current_snapshot if isinstance(current_snapshot, dict) else {},
+                report.wins, report.concerns,
+            )
+
+            # Target comparison
+            if targets_path:
+                try:
+                    targets_data = json.loads(Path(targets_path).read_text())
+                    # Use 30-day targets by default
+                    target_set = targets_data.get("30_day", targets_data)
+                    report.target_progress = self.compare_targets(
+                        current_snapshot if isinstance(current_snapshot, dict) else {},
+                        target_set,
+                    )
+                except Exception as exc:
+                    report.errors.append(f"Targets load error: {exc}")
+
+        return report
+
+
+# ---------------------------------------------------------------------------
+# Output formatting
+# ---------------------------------------------------------------------------
+
+def format_text_report(report: PerformanceReport) -> str:
+    """Format performance report as human-readable text."""
+    lines = []
+    lines.append("=" * 60)
+    lines.append(f"SEO Performance Report: {report.url}")
+    lines.append(f"Period: {report.period} ({report.date_from} to {report.date_to})")
+    lines.append(f"Generated: {report.timestamp}")
+    lines.append("=" * 60)
+
+    # Executive Summary
+    lines.append("\nEXECUTIVE SUMMARY")
+    lines.append("-" * 40)
+    es = report.executive_summary
+    lines.append(f"  Health Score: {es.get('health_score', 0)}/100")
+    trend_arrow = {"improving": "^", "declining": "v", "stable": "="}.get(
+        es.get("health_trend", "stable"), "="
+    )
+    lines.append(f"  Trend: {trend_arrow} {es.get('health_trend', 'stable')}")
+    lines.append(f"  Traffic Value Change: ${es.get('traffic_value_change_usd', 0):,.2f}")
+    lines.append(f"  Assessment: {es.get('overall_assessment', 'N/A')}")
+
+    # Wins
+    lines.append(f"\n  Top Wins ({es.get('total_wins', 0)} total):")
+    for w in es.get("top_wins", []):
+        impact_marker = {"high": "!!!", "medium": "!!", "low": "!"}.get(w.get("impact", "low"), "!")
+        lines.append(f"    {impact_marker} [{w.get('category', '')}] {w.get('description', '')}")
+
+    # Concerns
+    lines.append(f"\n  Top Concerns ({es.get('total_concerns', 0)} total):")
+    for c in es.get("top_concerns", []):
+        impact_marker = {"high": "!!!", "medium": "!!", "low": "!"}.get(c.get("impact", "low"), "!")
+        lines.append(f"    {impact_marker} [{c.get('category', '')}] {c.get('description', '')}")
+
+    # Trends
+    if report.trends:
+        lines.append("\nTRENDS")
+        lines.append("-" * 40)
+        for metric_name, trend_list in report.trends.items():
+            for t in trend_list:
+                if isinstance(t, dict):
+                    dir_arrow = {"up": "^", "down": "v", "stable": "="}.get(
+                        t.get("direction", "stable"), "="
+                    )
+                    change_str = f" ({t.get('change_pct', 0):+.1f}%)" if t.get("change_pct") is not None else ""
+                    lines.append(f"  {dir_arrow} {metric_name}: {t.get('value', 0):,.2f}{change_str}")
+
+    # Tactical Breakdown
+    if report.tactical_breakdown:
+        lines.append("\nTACTICAL BREAKDOWN")
+        lines.append("-" * 40)
+        for dim_name, dim_data in report.tactical_breakdown.items():
+            status = dim_data.get("status", "unknown")
+            status_marker = {
+                "improving": "^", "declining": "v", "active": "=", "needs_review": "?"
+            }.get(status, "?")
+            lines.append(f"\n  [{dim_name.upper()}] Status: {status_marker} {status}")
+            for action in dim_data.get("actions", [])[:3]:
+                lines.append(f"    > {action}")
+
+    # Target Progress
+    if report.target_progress:
+        lines.append("\nTARGET PROGRESS")
+        lines.append("-" * 40)
+        for tp in report.target_progress:
+            if isinstance(tp, TargetProgress):
+                bar_filled = int(min(tp.progress_pct, 100) / 5)
+                bar = "#" * bar_filled + "-" * (20 - bar_filled)
+                lines.append(
+                    f"  {tp.kpi_name}: [{bar}] {tp.progress_pct:.0f}% "
+                    f"(actual: {tp.actual:,.0f} / target: {tp.target:,.0f})"
+                )
+
+    # Errors
+    if report.errors:
+        lines.append("\nERRORS")
+        lines.append("-" * 40)
+        for err in report.errors:
+            lines.append(f"  ! {err}")
+
+    lines.append("\n" + "=" * 60)
+    return "\n".join(lines)
+
+
+def serialize_report(report: PerformanceReport) -> dict:
+    """Serialize PerformanceReport to JSON-safe dictionary."""
+    data = {
+        "url": report.url,
+        "period": report.period,
+        "date_from": report.date_from,
+        "date_to": report.date_to,
+        "health_score": report.health_score,
+        "health_trend": report.health_trend,
+        "trends": report.trends,
+        "wins": [asdict(w) for w in report.wins],
+        "concerns": [asdict(c) for c in report.concerns],
+        "executive_summary": report.executive_summary,
+        "tactical_breakdown": report.tactical_breakdown,
+        "target_progress": [asdict(tp) for tp in report.target_progress],
+        "traffic_value_change": report.traffic_value_change,
+        "timestamp": report.timestamp,
+        "errors": report.errors,
+    }
+    return data
+
+
+# ---------------------------------------------------------------------------
+# CLI entry point
+# ---------------------------------------------------------------------------
+
+def parse_args() -> argparse.Namespace:
+    """Parse command-line arguments."""
+    parser = argparse.ArgumentParser(
+        description="SEO Performance Reporter - Period-over-period analysis"
+    )
+    parser.add_argument(
+        "--url", required=True, help="Target URL or domain"
+    )
+    parser.add_argument(
+        "--period", choices=["monthly", "quarterly", "yearly", "custom"],
+        default="monthly", help="Report period (default: monthly)"
+    )
+    parser.add_argument(
+        "--from", dest="date_from", type=str, default=None,
+        help="Start date (YYYY-MM-DD) for custom period"
+    )
+    parser.add_argument(
+        "--to", dest="date_to", type=str, default=None,
+        help="End date (YYYY-MM-DD) for custom period"
+    )
+    parser.add_argument(
+        "--executive", action="store_true",
+        help="Generate executive summary only"
+    )
+    parser.add_argument(
+        "--targets", type=str, default=None,
+        help="Path to targets JSON file for progress comparison"
+    )
+    parser.add_argument(
+        "--json", action="store_true",
+        help="Output results as JSON"
+    )
+    parser.add_argument(
+        "--output", type=str, default=None,
+        help="Save output to file path"
+    )
+    return parser.parse_args()
+
+
+async def main() -> None:
+    """Main entry point."""
+    args = parse_args()
+
+    reporter = PerformanceReporter()
+    result = await reporter.report(
+        url=args.url,
+        period=args.period,
+        date_from=args.date_from,
+        date_to=args.date_to,
+        executive_only=args.executive,
+        targets_path=args.targets,
+    )
+
+    if args.json:
+        output = json.dumps(serialize_report(result), indent=2, ensure_ascii=False)
+    else:
+        output = format_text_report(result)
+
+    if args.output:
+        Path(args.output).write_text(output, encoding="utf-8")
+        logger.info(f"Output saved to {args.output}")
+    else:
+        print(output)
+
+    reporter.print_stats()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/25-seo-kpi-framework/code/scripts/requirements.txt b/custom-skills/25-seo-kpi-framework/code/scripts/requirements.txt
new file mode 100644
index 0000000..5c84aa3
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/code/scripts/requirements.txt
@@ -0,0 +1,8 @@
+# 25-seo-kpi-framework dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+pandas>=2.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/25-seo-kpi-framework/desktop/SKILL.md b/custom-skills/25-seo-kpi-framework/desktop/SKILL.md
new file mode 100644
index 0000000..846997e
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/desktop/SKILL.md
@@ -0,0 +1,107 @@
+---
+name: seo-kpi-framework
+description: |
+  SEO KPI and performance framework for unified metrics, health scores, ROI, and period-over-period reporting.
+  Triggers: SEO KPI, performance report, health score, SEO metrics, ROI, baseline, targets.
+---
+
+# SEO KPI & Performance Framework
+
+## Purpose
+
+Aggregate SEO KPIs across all dimensions into a unified dashboard. Establish baselines, set targets (30/60/90-day), generate executive summaries with health scores, provide tactical breakdowns, estimate ROI using Ahrefs traffic cost, and support period-over-period comparison (MoM, QoQ, YoY).
+
+## Core Capabilities
+
+1. **KPI Aggregation** - Unified metrics across 7 dimensions (traffic, rankings, links, technical, content, engagement, local)
+2. **Health Scoring** - Weighted 0-100 score with trend direction
+3. **Baseline & Targets** - Establish baselines and set 30/60/90 day growth targets
+4. **ROI Estimation** - Traffic value from Ahrefs organic cost
+5. **Performance Reporting** - Period-over-period comparison with executive summary
+6. **Tactical Breakdown** - Actionable next steps per dimension
+
+## MCP Tool Usage
+
+### Ahrefs for SEO Metrics
+```
+mcp__ahrefs__site-explorer-metrics: Current organic metrics snapshot
+mcp__ahrefs__site-explorer-metrics-history: Historical trend data
+mcp__ahrefs__site-explorer-metrics-by-country: Country-level breakdown
+mcp__ahrefs__site-explorer-domain-rating-history: Domain rating trend
+mcp__ahrefs__site-explorer-total-search-volume-history: Keyword volume trend
+```
+
+### Notion for Report Storage
+```
+mcp__notion__*: Save reports to SEO Audit Log database
+```
+
+## Workflow
+
+### 1. KPI Aggregation
+1. Fetch site-explorer-metrics for current organic data
+2. Extract traffic, ranking, link, technical, content metrics
+3. Calculate dimension scores with weights (traffic 25%, rankings 20%, technical 20%, content 15%, links 15%, local 5%)
+4. Compute overall health score (0-100)
+5. Set 30/60/90 day targets (5%/10%/20% improvement)
+6. Estimate ROI from Ahrefs traffic cost (divide raw cost by 100 for USD)
+
+### 2. Performance Reporting
+1. Determine date range from period (monthly/quarterly/yearly/custom)
+2. Fetch metrics-history for current and previous period
+3. Calculate period-over-period changes
+4. Identify wins (>5% improvement) and concerns (>5% decline)
+5. Generate executive summary with trend arrows
+6. Create tactical breakdown with actionable next steps
+7. Compare against targets if provided
+
+## Output Format
+
+```markdown
+## SEO KPI Dashboard: [domain]
+
+### Health Score: [score]/100 ([trend])
+
+### KPI Summary
+| Dimension | Score | Key Metric | Trend |
+|-----------|-------|------------|-------|
+| Traffic | [score] | [organic_traffic] | [arrow] |
+| Rankings | [score] | [visibility] | [arrow] |
+| Links | [score] | [DR] | [arrow] |
+| Technical | [score] | [health] | [arrow] |
+| Content | [score] | [indexed_pages] | [arrow] |
+
+### Executive Summary
+- Top Wins: [list]
+- Top Concerns: [list]
+- Recommendations: [list]
+
+### Targets (30/60/90 day)
+[Target table with progress bars]
+```
+
+## Key Metrics
+
+| Dimension | Metrics | Source |
+|-----------|---------|--------|
+| Traffic | Organic traffic, traffic value (USD) | site-explorer-metrics |
+| Rankings | Visibility score, top10 keywords | site-explorer-metrics |
+| Links | Domain rating, referring domains | domain-rating, metrics |
+| Technical | Pages crawled, technical health | site-explorer-metrics |
+| Content | Indexed pages, freshness score | site-explorer-metrics |
+| Local | GBP visibility, review score | External data |
+
+## Limitations
+
+- Local KPIs require external GBP data (not available via Ahrefs)
+- Engagement KPIs (bounce rate, session duration) require Google Analytics
+- Technical health is estimated heuristically from available data
+- ROI is estimated from Ahrefs traffic cost, not actual revenue
+
+## Notion Output (Required)
+
+All reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category, Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: KPI-YYYYMMDD-NNN
diff --git a/custom-skills/25-seo-kpi-framework/desktop/skill.yaml b/custom-skills/25-seo-kpi-framework/desktop/skill.yaml
new file mode 100644
index 0000000..a85a1e2
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-kpi-framework
+description: |
+  SEO KPI and performance framework. Triggers: SEO KPI, performance report, health score, SEO metrics, ROI, baseline, targets.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/25-seo-kpi-framework/desktop/tools/ahrefs.md b/custom-skills/25-seo-kpi-framework/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..5d918e0
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/desktop/tools/ahrefs.md
@@ -0,0 +1,15 @@
+# Ahrefs
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/25-seo-kpi-framework/desktop/tools/notion.md b/custom-skills/25-seo-kpi-framework/desktop/tools/notion.md
new file mode 100644
index 0000000..153e086
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/desktop/tools/notion.md
@@ -0,0 +1,15 @@
+# Notion
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/25-seo-kpi-framework/desktop/tools/websearch.md b/custom-skills/25-seo-kpi-framework/desktop/tools/websearch.md
new file mode 100644
index 0000000..36ef238
--- /dev/null
+++ b/custom-skills/25-seo-kpi-framework/desktop/tools/websearch.md
@@ -0,0 +1,15 @@
+# WebSearch
+
+> TODO: Document tool usage for this skill
+
+## Available Commands
+
+- [ ] List commands
+
+## Configuration
+
+- [ ] Add configuration details
+
+## Examples
+
+- [ ] Add usage examples
diff --git a/custom-skills/26-seo-international/code/CLAUDE.md b/custom-skills/26-seo-international/code/CLAUDE.md
new file mode 100644
index 0000000..e0a05ce
--- /dev/null
+++ b/custom-skills/26-seo-international/code/CLAUDE.md
@@ -0,0 +1,149 @@
+# CLAUDE.md
+
+## Overview
+
+International SEO audit tool for multi-language and multi-region website optimization. Validates hreflang tags (bidirectional, self-referencing, x-default), analyzes URL structure patterns (ccTLD vs subdomain vs subdirectory), audits content parity across language versions, checks language detection vs declared language, and analyzes international redirect logic. Supports Korean expansion patterns (ko→ja, ko→zh, ko→en).
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Hreflang validation
+python scripts/hreflang_validator.py --url https://example.com --json
+
+# Full international SEO audit
+python scripts/international_auditor.py --url https://example.com --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `hreflang_validator.py` | Validate hreflang tag implementation | Hreflang errors, missing bidirectional links, x-default issues |
+| `international_auditor.py` | Full international SEO audit | URL structure, content parity, redirect logic, language detection |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## Hreflang Validator
+
+```bash
+# Validate hreflang for homepage
+python scripts/hreflang_validator.py --url https://example.com --json
+
+# Validate with sitemap-based discovery
+python scripts/hreflang_validator.py --url https://example.com --sitemap https://example.com/sitemap.xml --json
+
+# Check specific pages
+python scripts/hreflang_validator.py --urls-file pages.txt --json
+```
+
+**Capabilities**:
+- Hreflang tag extraction from HTML head, HTTP headers, and XML sitemap
+- Bidirectional validation (if page A→B, then B→A must exist)
+- Self-referencing check (each page should reference itself)
+- x-default tag verification
+- Language/region code validation (ISO 639-1 + ISO 3166-1)
+- Conflicting hreflang detection
+- Missing language version detection
+- Return tag validation (confirmation links from alternate pages)
+
+## International Auditor
+
+```bash
+# Full international audit
+python scripts/international_auditor.py --url https://example.com --json
+
+# URL structure analysis
+python scripts/international_auditor.py --url https://example.com --scope structure --json
+
+# Content parity check
+python scripts/international_auditor.py --url https://example.com --scope parity --json
+
+# Korean expansion focus
+python scripts/international_auditor.py --url https://example.com --korean-expansion --json
+```
+
+**Capabilities**:
+- URL structure analysis (ccTLD vs subdomain vs subdirectory)
+  - Recommendation engine based on business context
+- Content parity audit across language versions
+  - Page count comparison per language
+  - Key page availability check (home, about, contact, products)
+  - Content freshness comparison across languages
+- Language/locale detection vs declared language
+  - HTML lang attribute check
+  - Content-Language header check
+  - Actual content language detection
+- International redirect logic audit
+  - IP-based redirect detection
+  - Accept-Language redirect behavior
+  - Geo-redirect best practices (suggest→don't force)
+- Korean expansion patterns (ko→ja, ko→zh, ko→en)
+  - Priority market recommendations for Korean businesses
+  - CJK-specific URL encoding issues
+  - Regional search engine considerations (Naver, Baidu, Yahoo Japan)
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `site-explorer-metrics-by-country` | Country-level traffic distribution |
+| `site-explorer-organic-keywords` | Keywords by country filter |
+
+## Output Format
+
+```json
+{
+  "url": "https://example.com",
+  "url_structure": "subdirectory",
+  "languages_detected": ["ko", "en", "ja"],
+  "hreflang_validation": {
+    "total_pages_checked": 50,
+    "errors": [],
+    "warnings": [],
+    "missing_bidirectional": [],
+    "missing_self_reference": [],
+    "x_default_present": true
+  },
+  "content_parity": {
+    "ko": {"pages": 150, "freshness_score": 90},
+    "en": {"pages": 120, "freshness_score": 75},
+    "ja": {"pages": 80, "freshness_score": 60}
+  },
+  "redirect_logic": {
+    "ip_based_redirect": false,
+    "language_based_redirect": true,
+    "is_forced": false
+  },
+  "score": 68,
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Audited website URL |
+| Category | Select | International SEO |
+| Priority | Select | Based on hreflang error count |
+| Found Date | Date | Audit date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: INTL-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., hreflang, x-default, ccTLD)
+- URLs and code remain unchanged
diff --git a/custom-skills/26-seo-international/code/scripts/base_client.py b/custom-skills/26-seo-international/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/26-seo-international/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/26-seo-international/code/scripts/international_auditor.py b/custom-skills/26-seo-international/code/scripts/international_auditor.py
new file mode 100644
index 0000000..15bcdcd
--- /dev/null
+++ b/custom-skills/26-seo-international/code/scripts/international_auditor.py
@@ -0,0 +1,1192 @@
+"""
+International SEO Auditor - Multi-language site audit
+=====================================================
+Purpose: Audit international SEO implementation including URL structure, content parity, language detection, and redirect logic
+Python: 3.10+
+Usage: python international_auditor.py --url https://example.com --json
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import re
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from typing import Optional
+from urllib.parse import urlparse, urljoin, parse_qs
+
+import aiohttp
+
+# ---------------------------------------------------------------------------
+# Logging
+# ---------------------------------------------------------------------------
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+logger = logging.getLogger("international_auditor")
+
+# ---------------------------------------------------------------------------
+# Constants
+# ---------------------------------------------------------------------------
+KEY_PAGES = ["home", "about", "contact", "products", "services", "blog", "faq"]
+
+CCTLD_MAP: dict[str, str] = {
+    ".kr": "ko", ".jp": "ja", ".cn": "zh", ".de": "de", ".fr": "fr",
+    ".es": "es", ".it": "it", ".pt": "pt", ".nl": "nl", ".ru": "ru",
+    ".br": "pt", ".mx": "es", ".co.uk": "en", ".ca": "en", ".au": "en",
+    ".in": "en", ".co.kr": "ko", ".co.jp": "ja", ".com.cn": "zh",
+}
+
+LANG_SUBDIRECTORY_PATTERN = re.compile(
+    r"^/(?P<lang>[a-z]{2}(?:-[a-z]{2})?)(?:/|$)", re.IGNORECASE,
+)
+
+LANG_SUBDOMAIN_PATTERN = re.compile(
+    r"^(?P<lang>[a-z]{2}(?:-[a-z]{2})?)\.", re.IGNORECASE,
+)
+
+ACCEPT_LANGUAGE_HEADERS = [
+    {"Accept-Language": "ko-KR,ko;q=0.9"},
+    {"Accept-Language": "en-US,en;q=0.9"},
+    {"Accept-Language": "ja-JP,ja;q=0.9"},
+    {"Accept-Language": "zh-CN,zh;q=0.9"},
+]
+
+URL_STRUCTURE_INFO: dict[str, dict] = {
+    "cctld": {
+        "pros": [
+            "Strong geo-targeting signal to search engines",
+            "Clear country association for users",
+            "Independent domain authority per country",
+        ],
+        "cons": [
+            "Expensive to acquire and maintain multiple domains",
+            "Domain authority is not shared across properties",
+            "Complex infrastructure management",
+        ],
+    },
+    "subdomain": {
+        "pros": [
+            "Easy to set up on a single domain",
+            "Can be hosted on different servers/locations",
+            "Clear separation of language versions",
+        ],
+        "cons": [
+            "Subdomains may be treated as separate sites by search engines",
+            "Limited geo-targeting signal compared to ccTLD",
+            "Domain authority dilution across subdomains",
+        ],
+    },
+    "subdirectory": {
+        "pros": [
+            "All content inherits root domain authority",
+            "Simplest infrastructure and lowest cost",
+            "Best for consolidating SEO signals",
+        ],
+        "cons": [
+            "Server must handle all language versions",
+            "Geo-targeting relies on Search Console settings and hreflang",
+            "Less clear country targeting than ccTLD",
+        ],
+    },
+    "parameter": {
+        "pros": [
+            "Easy to implement technically",
+        ],
+        "cons": [
+            "Not recommended by Google",
+            "URL parameters can be ignored by crawlers",
+            "Poor user experience and shareability",
+            "No geo-targeting signal",
+        ],
+    },
+}
+
+# ---------------------------------------------------------------------------
+# Unicode range helpers
+# ---------------------------------------------------------------------------
+RE_HANGUL = re.compile(r"[\uAC00-\uD7AF]")
+RE_HIRAGANA = re.compile(r"[\u3040-\u309F]")
+RE_KATAKANA = re.compile(r"[\u30A0-\u30FF]")
+RE_CJK_UNIFIED = re.compile(r"[\u4E00-\u9FFF]")
+RE_LATIN_EXTENDED = re.compile(r"[A-Za-z\u00C0-\u024F]")
+
+# ---------------------------------------------------------------------------
+# Dataclasses
+# ---------------------------------------------------------------------------
+
+@dataclass
+class URLStructureAnalysis:
+    """Analysis of the international URL structure pattern."""
+    pattern: str = ""                  # cctld / subdomain / subdirectory / parameter
+    examples: list[str] = field(default_factory=list)
+    languages_found: list[str] = field(default_factory=list)
+    recommendation: str = ""
+    pros: list[str] = field(default_factory=list)
+    cons: list[str] = field(default_factory=list)
+
+
+@dataclass
+class ContentParityEntry:
+    """Content parity data for a single language version."""
+    language: str = ""
+    page_count: int = 0
+    key_pages_present: list[str] = field(default_factory=list)
+    freshness_score: float = 0.0
+    missing_pages: list[str] = field(default_factory=list)
+
+
+@dataclass
+class LanguageDetection:
+    """Language detection audit for a single page."""
+    page_url: str = ""
+    declared_lang: str = ""
+    html_lang_attr: str = ""
+    content_language_header: str = ""
+    detected_lang: str = ""
+    is_consistent: bool = True
+
+
+@dataclass
+class RedirectLogic:
+    """International redirect behaviour audit."""
+    has_ip_redirect: bool = False
+    has_language_redirect: bool = False
+    is_forced: bool = False
+    redirect_type: str = ""            # 301 / 302 / js / none
+    recommendation: str = ""
+
+
+@dataclass
+class KoreanExpansion:
+    """Korean-market expansion analysis."""
+    primary_lang: str = "ko"
+    target_markets: list[str] = field(default_factory=list)
+    cjk_url_issues: list[str] = field(default_factory=list)
+    regional_search_engines: dict[str, str] = field(default_factory=dict)
+    priority_recommendations: list[str] = field(default_factory=list)
+
+
+@dataclass
+class InternationalAuditResult:
+    """Top-level result container for the full international audit."""
+    url: str = ""
+    url_structure: Optional[URLStructureAnalysis] = None
+    languages_detected: list[str] = field(default_factory=list)
+    content_parity: dict[str, ContentParityEntry] = field(default_factory=dict)
+    language_detection_issues: list[LanguageDetection] = field(default_factory=list)
+    redirect_logic: Optional[RedirectLogic] = None
+    korean_expansion: Optional[KoreanExpansion] = None
+    issues: list[str] = field(default_factory=list)
+    recommendations: list[str] = field(default_factory=list)
+    score: int = 100
+    timestamp: str = ""
+
+
+# ---------------------------------------------------------------------------
+# Auditor
+# ---------------------------------------------------------------------------
+
+class InternationalAuditor:
+    """Perform a comprehensive international SEO audit on a target URL."""
+
+    def __init__(
+        self,
+        url: str,
+        korean_expansion: bool = False,
+        scope: str | None = None,
+    ):
+        self.url = url.rstrip("/")
+        self.korean_expansion_flag = korean_expansion
+        self.scope = scope or "all"
+        self.parsed = urlparse(self.url)
+        self.base_url = f"{self.parsed.scheme}://{self.parsed.netloc}"
+        self.timeout = aiohttp.ClientTimeout(total=30)
+        self._session: aiohttp.ClientSession | None = None
+
+    # ------------------------------------------------------------------
+    # Public entry point
+    # ------------------------------------------------------------------
+
+    async def audit(self) -> InternationalAuditResult:
+        """Run the full international SEO audit and return results."""
+        result = InternationalAuditResult(
+            url=self.url,
+            timestamp=datetime.utcnow().isoformat(),
+        )
+
+        async with aiohttp.ClientSession(timeout=self.timeout) as session:
+            self._session = session
+
+            try:
+                # 1. URL structure analysis
+                if self.scope in ("all", "structure"):
+                    logger.info("Analysing URL structure ...")
+                    result.url_structure = await self._analyze_url_structure(self.url)
+
+                # 2. Discover language versions
+                logger.info("Discovering language versions ...")
+                lang_versions = await self._discover_language_versions(self.url)
+                result.languages_detected = list(lang_versions.keys())
+
+                # 3. Content parity
+                if self.scope in ("all", "parity"):
+                    logger.info("Auditing content parity ...")
+                    result.content_parity = await self._audit_content_parity(lang_versions)
+
+                # 4. Language detection
+                if self.scope in ("all", "detection"):
+                    logger.info("Checking language detection consistency ...")
+                    all_pages: list[tuple[str, str]] = []
+                    for lang, urls in lang_versions.items():
+                        for u in urls[:5]:
+                            all_pages.append((lang, u))
+                    detections = await self._check_language_detection(all_pages)
+                    result.language_detection_issues = [
+                        d for d in detections if not d.is_consistent
+                    ]
+
+                # 5. Redirect logic
+                if self.scope in ("all", "redirects"):
+                    logger.info("Auditing redirect logic ...")
+                    result.redirect_logic = await self._audit_redirect_logic(self.url)
+
+                # 6. Korean expansion
+                if self.korean_expansion_flag:
+                    logger.info("Analysing Korean expansion opportunities ...")
+                    result.korean_expansion = await self._analyze_korean_expansion(self.url)
+
+                # Collect issues and recommendations
+                self._collect_issues(result)
+                result.score = self._calculate_score(result)
+
+            except Exception as exc:
+                logger.error("Audit failed: %s", exc)
+                result.issues.append(f"Audit error: {exc}")
+                result.score = 0
+
+            finally:
+                self._session = None
+
+        return result
+
+    # ------------------------------------------------------------------
+    # URL structure analysis
+    # ------------------------------------------------------------------
+
+    async def _analyze_url_structure(self, url: str) -> URLStructureAnalysis:
+        """Determine the international URL pattern in use."""
+        parsed = urlparse(url)
+        hostname = parsed.netloc.lower()
+        path = parsed.path
+
+        analysis = URLStructureAnalysis()
+
+        # Check ccTLD
+        for tld, lang in CCTLD_MAP.items():
+            if hostname.endswith(tld):
+                analysis.pattern = "cctld"
+                analysis.examples.append(f"{hostname} -> {lang}")
+                analysis.languages_found.append(lang)
+                break
+
+        # Check subdomain pattern
+        if not analysis.pattern:
+            match = LANG_SUBDOMAIN_PATTERN.match(hostname)
+            if match:
+                lang_code = match.group("lang").lower()
+                if len(lang_code) == 2 and lang_code not in ("ww", "ww"):
+                    analysis.pattern = "subdomain"
+                    analysis.examples.append(f"{hostname}")
+                    analysis.languages_found.append(lang_code)
+
+        # Check subdirectory pattern
+        if not analysis.pattern:
+            match = LANG_SUBDIRECTORY_PATTERN.match(path)
+            if match:
+                lang_code = match.group("lang").lower()
+                analysis.pattern = "subdirectory"
+                analysis.examples.append(f"{path}")
+                analysis.languages_found.append(lang_code)
+
+        # Check URL parameter
+        if not analysis.pattern:
+            qs = parse_qs(parsed.query)
+            for param in ("lang", "language", "hl", "locale", "lg"):
+                if param in qs:
+                    analysis.pattern = "parameter"
+                    analysis.examples.append(f"?{param}={qs[param][0]}")
+                    analysis.languages_found.append(qs[param][0])
+                    break
+
+        # Try to discover additional language versions from page HTML
+        try:
+            html = await self._fetch_text(url)
+            if html:
+                hreflang_langs = self._extract_hreflang_langs(html)
+                for lang in hreflang_langs:
+                    if lang not in analysis.languages_found:
+                        analysis.languages_found.append(lang)
+
+                # Refine pattern from hreflang hrefs if not yet determined
+                if not analysis.pattern and hreflang_langs:
+                    hrefs = self._extract_hreflang_hrefs(html)
+                    analysis.pattern = self._detect_pattern_from_hrefs(hrefs)
+        except Exception as exc:
+            logger.warning("Could not fetch page for structure analysis: %s", exc)
+
+        if not analysis.pattern:
+            analysis.pattern = "unknown"
+
+        # Attach pros/cons/recommendation
+        info = URL_STRUCTURE_INFO.get(analysis.pattern, {})
+        analysis.pros = info.get("pros", [])
+        analysis.cons = info.get("cons", [])
+
+        if analysis.pattern == "parameter":
+            analysis.recommendation = (
+                "URL parameters are not recommended for international targeting. "
+                "Consider migrating to subdirectories or subdomains."
+            )
+        elif analysis.pattern == "cctld":
+            analysis.recommendation = (
+                "ccTLDs provide the strongest geo-targeting signal. "
+                "Ensure each domain has sufficient content and backlinks."
+            )
+        elif analysis.pattern == "subdomain":
+            analysis.recommendation = (
+                "Subdomains can work well but ensure hreflang tags are properly "
+                "implemented to signal relationships between language versions."
+            )
+        elif analysis.pattern == "subdirectory":
+            analysis.recommendation = (
+                "Subdirectories are the most cost-effective approach and consolidate "
+                "domain authority. This is the recommended pattern for most sites."
+            )
+        else:
+            analysis.recommendation = (
+                "Could not determine the URL structure pattern. "
+                "Implement a clear international URL strategy using subdirectories, "
+                "subdomains, or ccTLDs with proper hreflang tags."
+            )
+
+        return analysis
+
+    # ------------------------------------------------------------------
+    # Discover language versions
+    # ------------------------------------------------------------------
+
+    async def _discover_language_versions(self, url: str) -> dict[str, list[str]]:
+        """Discover language versions from hreflang tags and common patterns."""
+        lang_versions: dict[str, list[str]] = {}
+
+        try:
+            html = await self._fetch_text(url)
+        except Exception:
+            html = None
+
+        # 1. Extract from hreflang tags
+        if html:
+            hreflang_map = self._extract_hreflang_map(html)
+            for lang, href in hreflang_map.items():
+                if lang == "x-default":
+                    continue
+                lang_key = lang.split("-")[0].lower()
+                lang_versions.setdefault(lang_key, []).append(href)
+
+        # 2. Probe common subdirectory patterns
+        common_langs = ["en", "ko", "ja", "zh", "de", "fr", "es", "pt"]
+        probe_tasks = []
+        for lang in common_langs:
+            if lang not in lang_versions:
+                candidate = f"{self.base_url}/{lang}/"
+                probe_tasks.append(self._probe_url(lang, candidate))
+
+        if probe_tasks:
+            results = await asyncio.gather(*probe_tasks, return_exceptions=True)
+            for item in results:
+                if isinstance(item, tuple):
+                    lang, probe_url = item
+                    lang_versions.setdefault(lang, []).append(probe_url)
+
+        # 3. Probe common subdomain patterns
+        domain_parts = self.parsed.netloc.split(".")
+        if len(domain_parts) >= 2:
+            root_domain = ".".join(domain_parts[-2:])
+            subdomain_tasks = []
+            for lang in common_langs:
+                if lang not in lang_versions:
+                    candidate = f"{self.parsed.scheme}://{lang}.{root_domain}/"
+                    subdomain_tasks.append(self._probe_url(lang, candidate))
+            if subdomain_tasks:
+                results = await asyncio.gather(*subdomain_tasks, return_exceptions=True)
+                for item in results:
+                    if isinstance(item, tuple):
+                        lang, probe_url = item
+                        lang_versions.setdefault(lang, []).append(probe_url)
+
+        # Ensure the original URL is listed under some language
+        if not lang_versions:
+            detected = "unknown"
+            if html:
+                detected = self._detect_language_from_content(html) or "unknown"
+            lang_versions[detected] = [url]
+
+        return lang_versions
+
+    async def _probe_url(self, lang: str, url: str) -> tuple[str, str] | None:
+        """Check if a URL returns 200 and return (lang, url) or raise."""
+        try:
+            async with self._session.head(
+                url, allow_redirects=True, timeout=self.timeout,
+            ) as resp:
+                if resp.status == 200:
+                    return (lang, url)
+        except Exception:
+            pass
+        raise ValueError(f"Probe failed for {url}")
+
+    # ------------------------------------------------------------------
+    # Content parity
+    # ------------------------------------------------------------------
+
+    async def _audit_content_parity(
+        self,
+        lang_versions: dict[str, list[str]],
+    ) -> dict[str, ContentParityEntry]:
+        """Audit content parity across discovered language versions."""
+        parity: dict[str, ContentParityEntry] = {}
+
+        for lang, urls in lang_versions.items():
+            entry = ContentParityEntry(language=lang, page_count=len(urls))
+
+            # Determine base URL for this language version
+            if urls:
+                lang_base = urls[0].rstrip("/")
+                present_keys = await self._check_key_pages(lang_base)
+                entry.key_pages_present = present_keys
+                entry.missing_pages = [
+                    p for p in KEY_PAGES if p not in present_keys
+                ]
+
+                # Freshness score from Last-Modified header on first URL
+                entry.freshness_score = await self._get_freshness_score(urls[0])
+
+            parity[lang] = entry
+
+        return parity
+
+    async def _check_key_pages(self, lang_base_url: str) -> list[str]:
+        """Check which key pages exist under a language base URL."""
+        present: list[str] = []
+        page_paths = {
+            "home": "/",
+            "about": "/about",
+            "contact": "/contact",
+            "products": "/products",
+            "services": "/services",
+            "blog": "/blog",
+            "faq": "/faq",
+        }
+        tasks = []
+        for name, path in page_paths.items():
+            full_url = lang_base_url.rstrip("/") + path
+            tasks.append(self._check_page_exists(name, full_url))
+
+        results = await asyncio.gather(*tasks, return_exceptions=True)
+        for item in results:
+            if isinstance(item, str):
+                present.append(item)
+        return present
+
+    async def _check_page_exists(self, name: str, url: str) -> str | None:
+        """Return the page name if URL returns 200, else raise."""
+        try:
+            async with self._session.head(
+                url, allow_redirects=True, timeout=self.timeout,
+            ) as resp:
+                if resp.status == 200:
+                    return name
+        except Exception:
+            pass
+        raise ValueError(f"Page not found: {url}")
+
+    async def _get_freshness_score(self, url: str) -> float:
+        """Compute a freshness score (0-100) based on Last-Modified header."""
+        try:
+            async with self._session.head(
+                url, allow_redirects=True, timeout=self.timeout,
+            ) as resp:
+                last_modified = resp.headers.get("Last-Modified")
+                if last_modified:
+                    from email.utils import parsedate_to_datetime
+                    mod_dt = parsedate_to_datetime(last_modified)
+                    age_days = (datetime.utcnow() - mod_dt.replace(tzinfo=None)).days
+                    if age_days <= 7:
+                        return 100.0
+                    elif age_days <= 30:
+                        return 90.0
+                    elif age_days <= 90:
+                        return 70.0
+                    elif age_days <= 180:
+                        return 50.0
+                    elif age_days <= 365:
+                        return 30.0
+                    else:
+                        return 10.0
+                # No Last-Modified header -- neutral score
+                return 50.0
+        except Exception:
+            return 0.0
+
+    # ------------------------------------------------------------------
+    # Language detection
+    # ------------------------------------------------------------------
+
+    async def _check_language_detection(
+        self,
+        pages: list[tuple[str, str]],
+    ) -> list[LanguageDetection]:
+        """Check language declaration consistency for a list of pages."""
+        tasks = [
+            self._detect_single_page(declared_lang, page_url)
+            for declared_lang, page_url in pages
+        ]
+        results = await asyncio.gather(*tasks, return_exceptions=True)
+        detections: list[LanguageDetection] = []
+        for r in results:
+            if isinstance(r, LanguageDetection):
+                detections.append(r)
+        return detections
+
+    async def _detect_single_page(
+        self,
+        declared_lang: str,
+        page_url: str,
+    ) -> LanguageDetection:
+        """Perform language detection checks on a single page."""
+        detection = LanguageDetection(
+            page_url=page_url,
+            declared_lang=declared_lang,
+        )
+
+        try:
+            async with self._session.get(
+                page_url, allow_redirects=True, timeout=self.timeout,
+            ) as resp:
+                # Content-Language header
+                cl_header = resp.headers.get("Content-Language", "")
+                detection.content_language_header = cl_header.strip()
+
+                html = await resp.text(errors="replace")
+
+                # HTML lang attribute
+                lang_match = re.search(
+                    r'<html[^>]*\slang=["\']([^"\']+)["\']',
+                    html,
+                    re.IGNORECASE,
+                )
+                if lang_match:
+                    detection.html_lang_attr = lang_match.group(1).strip()
+
+                # Detect from content
+                detection.detected_lang = self._detect_language_from_content(html)
+
+                # Consistency check
+                expected = declared_lang.split("-")[0].lower()
+                html_lang_base = detection.html_lang_attr.split("-")[0].lower() if detection.html_lang_attr else ""
+                cl_base = detection.content_language_header.split("-")[0].lower() if detection.content_language_header else ""
+                detected_base = detection.detected_lang.split("-")[0].lower() if detection.detected_lang else ""
+
+                inconsistencies = []
+                if html_lang_base and html_lang_base != expected:
+                    inconsistencies.append("html_lang_attr")
+                if cl_base and cl_base != expected:
+                    inconsistencies.append("content_language_header")
+                if detected_base and detected_base != expected:
+                    inconsistencies.append("detected_content")
+
+                detection.is_consistent = len(inconsistencies) == 0
+
+        except Exception as exc:
+            logger.warning("Language detection failed for %s: %s", page_url, exc)
+            detection.is_consistent = False
+
+        return detection
+
+    def _detect_language_from_content(self, html: str) -> str:
+        """Simple heuristic language detection using Unicode ranges."""
+        # Strip HTML tags for content analysis
+        text = re.sub(r"<[^>]+>", " ", html)
+        text = re.sub(r"\s+", " ", text)[:5000]  # limit analysis scope
+
+        counts = {
+            "ko": len(RE_HANGUL.findall(text)),
+            "ja": len(RE_HIRAGANA.findall(text)) + len(RE_KATAKANA.findall(text)),
+            "zh": len(RE_CJK_UNIFIED.findall(text)),
+            "en": len(RE_LATIN_EXTENDED.findall(text)),
+        }
+
+        # Japanese text also contains CJK unified characters (kanji).
+        # If both hiragana/katakana AND CJK are present, prefer Japanese.
+        if counts["ja"] > 0 and counts["zh"] > 0:
+            if counts["ja"] > counts["zh"] * 0.1:
+                counts["zh"] = 0  # attribute CJK chars to Japanese context
+
+        # Korean text can contain some CJK (hanja) but Hangul dominates.
+        if counts["ko"] > 0 and counts["zh"] > 0:
+            if counts["ko"] > counts["zh"]:
+                counts["zh"] = 0
+
+        if not any(counts.values()):
+            return "unknown"
+
+        dominant = max(counts, key=counts.get)
+
+        # Require a meaningful amount of the dominant script
+        if counts[dominant] < 20:
+            return "unknown"
+
+        return dominant
+
+    # ------------------------------------------------------------------
+    # Redirect logic
+    # ------------------------------------------------------------------
+
+    async def _audit_redirect_logic(self, url: str) -> RedirectLogic:
+        """Audit redirect behaviour for different Accept-Language headers."""
+        logic = RedirectLogic()
+        redirect_targets: dict[str, str] = {}
+
+        for headers in ACCEPT_LANGUAGE_HEADERS:
+            try:
+                async with self._session.get(
+                    url,
+                    allow_redirects=False,
+                    headers=headers,
+                    timeout=self.timeout,
+                ) as resp:
+                    lang_key = list(headers.values())[0].split(",")[0]
+                    if resp.status in (301, 302, 303, 307, 308):
+                        location = resp.headers.get("Location", "")
+                        redirect_targets[lang_key] = location
+                        logic.has_language_redirect = True
+                        if resp.status in (301, 302):
+                            logic.redirect_type = str(resp.status)
+                            logic.is_forced = True
+                    else:
+                        redirect_targets[lang_key] = str(resp.url)
+            except Exception as exc:
+                logger.warning("Redirect check failed for %s: %s", headers, exc)
+
+        # Detect IP-based redirects by comparing redirect targets
+        unique_targets = set(redirect_targets.values())
+        if len(unique_targets) > 1:
+            logic.has_language_redirect = True
+
+        # Check for JS-based redirects in page source
+        try:
+            html = await self._fetch_text(url)
+            if html:
+                js_redirect_patterns = [
+                    r"navigator\.language",
+                    r"navigator\.languages",
+                    r"window\.location\s*=",
+                    r"location\.href\s*=",
+                    r"geo[_-]?redirect",
+                    r"ip[_-]?redirect",
+                ]
+                for pattern in js_redirect_patterns:
+                    if re.search(pattern, html, re.IGNORECASE):
+                        if "geo" in pattern or "ip" in pattern:
+                            logic.has_ip_redirect = True
+                        if not logic.redirect_type:
+                            logic.redirect_type = "js"
+                        break
+        except Exception:
+            pass
+
+        # Recommendation
+        if logic.is_forced:
+            logic.recommendation = (
+                "Forced language/region redirects detected (HTTP {}). "
+                "Best practice: suggest language via a banner or interstitial "
+                "rather than forcing a redirect. Forced redirects can prevent "
+                "search engines from crawling all language versions and frustrate "
+                "users who prefer a different language."
+            ).format(logic.redirect_type)
+        elif logic.has_language_redirect:
+            logic.recommendation = (
+                "Language-based content negotiation detected. Ensure all language "
+                "versions remain directly accessible via their canonical URLs and "
+                "that Googlebot can crawl every version without redirect loops."
+            )
+        else:
+            logic.recommendation = (
+                "No forced language redirects detected. Consider adding a "
+                "language suggestion banner for users whose browser language "
+                "differs from the page language."
+            )
+
+        return logic
+
+    # ------------------------------------------------------------------
+    # Korean expansion analysis
+    # ------------------------------------------------------------------
+
+    async def _analyze_korean_expansion(self, url: str) -> KoreanExpansion:
+        """Analyse expansion opportunities for Korean-primary sites."""
+        expansion = KoreanExpansion(
+            primary_lang="ko",
+            target_markets=["ja", "zh", "en"],
+            regional_search_engines={
+                "ko": "Naver (naver.com) -- dominant in Korea; also consider Daum",
+                "ja": "Yahoo Japan (yahoo.co.jp) -- significant market share alongside Google",
+                "zh": "Baidu (baidu.com) -- dominant in mainland China; also Sogou, 360 Search",
+                "en": "Google (google.com) -- global default",
+            },
+        )
+
+        # Check CJK URL encoding issues
+        try:
+            html = await self._fetch_text(url)
+            if html:
+                # Find links with non-ASCII characters in URLs
+                href_pattern = re.compile(r'href=["\']([^"\']+)["\']', re.IGNORECASE)
+                for match in href_pattern.finditer(html):
+                    href = match.group(1)
+                    # Check for unencoded CJK characters in URLs
+                    if RE_HANGUL.search(href):
+                        expansion.cjk_url_issues.append(
+                            f"Unencoded Korean characters in URL: {href[:100]}"
+                        )
+                    if RE_CJK_UNIFIED.search(href):
+                        expansion.cjk_url_issues.append(
+                            f"Unencoded CJK characters in URL: {href[:100]}"
+                        )
+                    if RE_HIRAGANA.search(href) or RE_KATAKANA.search(href):
+                        expansion.cjk_url_issues.append(
+                            f"Unencoded Japanese characters in URL: {href[:100]}"
+                        )
+
+                # De-duplicate
+                expansion.cjk_url_issues = list(dict.fromkeys(expansion.cjk_url_issues))[:20]
+        except Exception as exc:
+            logger.warning("CJK URL check failed: %s", exc)
+
+        # Priority recommendations
+        expansion.priority_recommendations = [
+            "Implement hreflang tags linking ko, ja, zh, and en versions with x-default",
+            "Use subdirectories (/ko/, /ja/, /zh/, /en/) for cost-effective international structure",
+            "Create a Naver Search Advisor (searchadvisor.naver.com) account for Korean SEO",
+            "Submit sitemap to Baidu Webmaster Tools for Chinese market visibility",
+            "Register with Yahoo Japan Search Console (optional but recommended for Japan)",
+            "Ensure all URLs use percent-encoded paths -- avoid raw CJK characters in hrefs",
+            "Provide a language/region selector accessible from every page",
+            "Localise content beyond translation: currency, date formats, cultural references",
+            "Consider separate social media strategies per market (KakaoTalk for KR, LINE for JP, WeChat for CN)",
+            "Monitor Core Web Vitals per region -- CDN edge presence matters for CJK markets",
+        ]
+
+        return expansion
+
+    # ------------------------------------------------------------------
+    # Issue collection
+    # ------------------------------------------------------------------
+
+    def _collect_issues(self, result: InternationalAuditResult) -> None:
+        """Populate issues and recommendations from audit sub-results."""
+        # URL structure issues
+        if result.url_structure:
+            if result.url_structure.pattern == "parameter":
+                result.issues.append(
+                    "URL parameters used for language targeting -- not recommended"
+                )
+            if result.url_structure.pattern == "unknown":
+                result.issues.append(
+                    "Could not detect a clear international URL structure"
+                )
+            if len(result.url_structure.languages_found) <= 1:
+                result.issues.append(
+                    "Only one language version detected -- consider expanding"
+                )
+
+        # Content parity issues
+        if result.content_parity:
+            primary_count = max(
+                (e.page_count for e in result.content_parity.values()), default=0,
+            )
+            for lang, entry in result.content_parity.items():
+                if entry.missing_pages:
+                    result.issues.append(
+                        f"Language '{lang}' is missing key pages: "
+                        f"{', '.join(entry.missing_pages)}"
+                    )
+                if primary_count > 0 and entry.page_count < primary_count * 0.5:
+                    result.issues.append(
+                        f"Language '{lang}' has significantly fewer pages "
+                        f"({entry.page_count}) than the primary version ({primary_count})"
+                    )
+
+        # Language detection issues
+        for det in result.language_detection_issues:
+            result.issues.append(
+                f"Language inconsistency on {det.page_url}: "
+                f"declared={det.declared_lang}, html_lang={det.html_lang_attr}, "
+                f"detected={det.detected_lang}"
+            )
+
+        # Redirect logic issues
+        if result.redirect_logic:
+            if result.redirect_logic.is_forced:
+                result.issues.append(
+                    "Forced language/region redirects detected -- "
+                    "this can block search engine crawling"
+                )
+            if result.redirect_logic.has_ip_redirect:
+                result.issues.append(
+                    "IP-based redirect logic detected -- "
+                    "ensure Googlebot can access all versions"
+                )
+
+        # Check for x-default in hreflang
+        has_xdefault = False
+        if result.url_structure and result.url_structure.languages_found:
+            try:
+                html = None  # already fetched; recheck from structure data
+                # We check for x-default in languages_found (hreflang extraction adds it)
+                if "x-default" in result.url_structure.languages_found:
+                    has_xdefault = True
+                    result.url_structure.languages_found.remove("x-default")
+            except Exception:
+                pass
+        if not has_xdefault and len(result.languages_detected) > 1:
+            result.issues.append(
+                "No x-default hreflang tag detected -- "
+                "add x-default to specify the fallback page"
+            )
+
+        # Recommendations
+        if result.issues:
+            result.recommendations.append(
+                "Address the issues listed above to improve international SEO health"
+            )
+        if result.url_structure and result.url_structure.recommendation:
+            result.recommendations.append(result.url_structure.recommendation)
+        if result.redirect_logic and result.redirect_logic.recommendation:
+            result.recommendations.append(result.redirect_logic.recommendation)
+        if result.korean_expansion:
+            result.recommendations.extend(
+                result.korean_expansion.priority_recommendations[:5]
+            )
+
+    # ------------------------------------------------------------------
+    # Score calculation
+    # ------------------------------------------------------------------
+
+    def _calculate_score(self, result: InternationalAuditResult) -> int:
+        """Calculate an overall international SEO score (0-100)."""
+        score = 100
+
+        # URL structure penalties
+        if result.url_structure:
+            if result.url_structure.pattern == "parameter":
+                score -= 15
+            elif result.url_structure.pattern == "unknown":
+                score -= 10
+            if len(result.url_structure.languages_found) <= 1:
+                score -= 5
+
+        # Content parity penalties
+        if result.content_parity:
+            for entry in result.content_parity.values():
+                score -= 5 * len(entry.missing_pages)
+
+        # Language detection penalties
+        score -= 10 * len(result.language_detection_issues)
+
+        # Redirect logic penalties
+        if result.redirect_logic:
+            if result.redirect_logic.is_forced:
+                score -= 15
+
+        # No x-default penalty
+        for issue in result.issues:
+            if "x-default" in issue:
+                score -= 10
+                break
+
+        # CJK encoding issues
+        if result.korean_expansion:
+            cjk_count = len(result.korean_expansion.cjk_url_issues)
+            score -= 5 * min(cjk_count, 5)  # cap at 25 points
+
+        return max(0, score)
+
+    # ------------------------------------------------------------------
+    # HTML / hreflang helpers
+    # ------------------------------------------------------------------
+
+    async def _fetch_text(self, url: str) -> str | None:
+        """Fetch URL content as text."""
+        try:
+            async with self._session.get(
+                url, allow_redirects=True, timeout=self.timeout,
+            ) as resp:
+                if resp.status == 200:
+                    return await resp.text(errors="replace")
+        except Exception as exc:
+            logger.warning("Failed to fetch %s: %s", url, exc)
+        return None
+
+    def _extract_hreflang_langs(self, html: str) -> list[str]:
+        """Extract language codes from hreflang link tags."""
+        pattern = re.compile(
+            r'<link[^>]*\brel=["\']alternate["\'][^>]*\bhreflang=["\']([^"\']+)["\']',
+            re.IGNORECASE,
+        )
+        return [m.group(1).lower() for m in pattern.finditer(html)]
+
+    def _extract_hreflang_hrefs(self, html: str) -> list[str]:
+        """Extract href values from hreflang link tags."""
+        pattern = re.compile(
+            r'<link[^>]*\brel=["\']alternate["\'][^>]*\bhref=["\']([^"\']+)["\']',
+            re.IGNORECASE,
+        )
+        return [m.group(1) for m in pattern.finditer(html)]
+
+    def _extract_hreflang_map(self, html: str) -> dict[str, str]:
+        """Extract a mapping of hreflang language code to href."""
+        result: dict[str, str] = {}
+        # Match both attribute orderings
+        pattern1 = re.compile(
+            r'<link[^>]*\brel=["\']alternate["\']'
+            r'[^>]*\bhreflang=["\']([^"\']+)["\']'
+            r'[^>]*\bhref=["\']([^"\']+)["\']',
+            re.IGNORECASE,
+        )
+        pattern2 = re.compile(
+            r'<link[^>]*\bhreflang=["\']([^"\']+)["\']'
+            r'[^>]*\bhref=["\']([^"\']+)["\']'
+            r'[^>]*\brel=["\']alternate["\']',
+            re.IGNORECASE,
+        )
+        for m in pattern1.finditer(html):
+            result[m.group(1).lower()] = m.group(2)
+        for m in pattern2.finditer(html):
+            result[m.group(1).lower()] = m.group(2)
+        return result
+
+    def _detect_pattern_from_hrefs(self, hrefs: list[str]) -> str:
+        """Guess the URL structure pattern from a list of hreflang hrefs."""
+        for href in hrefs:
+            parsed = urlparse(href)
+            # ccTLD check
+            for tld in CCTLD_MAP:
+                if parsed.netloc.endswith(tld):
+                    return "cctld"
+            # subdomain check
+            if LANG_SUBDOMAIN_PATTERN.match(parsed.netloc):
+                return "subdomain"
+            # subdirectory check
+            if LANG_SUBDIRECTORY_PATTERN.match(parsed.path):
+                return "subdirectory"
+            # parameter check
+            qs = parse_qs(parsed.query)
+            for param in ("lang", "language", "hl", "locale", "lg"):
+                if param in qs:
+                    return "parameter"
+        return "unknown"
+
+
+# ---------------------------------------------------------------------------
+# Serialisation helper
+# ---------------------------------------------------------------------------
+
+def _result_to_dict(result: InternationalAuditResult) -> dict:
+    """Convert the audit result dataclass tree to a plain dict."""
+    data = asdict(result)
+    # Flatten content_parity keys
+    if "content_parity" in data and isinstance(data["content_parity"], dict):
+        cleaned = {}
+        for key, value in data["content_parity"].items():
+            cleaned[key] = value
+        data["content_parity"] = cleaned
+    return data
+
+
+def _print_text_report(result: InternationalAuditResult) -> None:
+    """Print a human-readable text report to stdout."""
+    print("=" * 70)
+    print("  International SEO Audit Report")
+    print("=" * 70)
+    print(f"  URL:       {result.url}")
+    print(f"  Timestamp: {result.timestamp}")
+    print(f"  Score:     {result.score}/100")
+    print("=" * 70)
+
+    # URL Structure
+    if result.url_structure:
+        print("\n--- URL Structure ---")
+        print(f"  Pattern:    {result.url_structure.pattern}")
+        print(f"  Languages:  {', '.join(result.url_structure.languages_found)}")
+        if result.url_structure.examples:
+            print(f"  Examples:   {', '.join(result.url_structure.examples)}")
+        print(f"  Recommendation: {result.url_structure.recommendation}")
+
+    # Languages detected
+    if result.languages_detected:
+        print(f"\n--- Languages Detected ---")
+        print(f"  {', '.join(result.languages_detected)}")
+
+    # Content parity
+    if result.content_parity:
+        print("\n--- Content Parity ---")
+        for lang, entry in result.content_parity.items():
+            missing = ", ".join(entry.missing_pages) if entry.missing_pages else "none"
+            present = ", ".join(entry.key_pages_present) if entry.key_pages_present else "none"
+            print(f"  [{lang}] pages={entry.page_count}  freshness={entry.freshness_score:.0f}")
+            print(f"         key pages present: {present}")
+            print(f"         missing: {missing}")
+
+    # Language detection issues
+    if result.language_detection_issues:
+        print("\n--- Language Detection Issues ---")
+        for det in result.language_detection_issues:
+            print(f"  {det.page_url}")
+            print(f"    declared={det.declared_lang}  html_lang={det.html_lang_attr}  "
+                  f"header={det.content_language_header}  detected={det.detected_lang}")
+
+    # Redirect logic
+    if result.redirect_logic:
+        print("\n--- Redirect Logic ---")
+        print(f"  Language redirect: {result.redirect_logic.has_language_redirect}")
+        print(f"  IP redirect:      {result.redirect_logic.has_ip_redirect}")
+        print(f"  Forced:           {result.redirect_logic.is_forced}")
+        print(f"  Type:             {result.redirect_logic.redirect_type or 'none'}")
+        print(f"  Recommendation:   {result.redirect_logic.recommendation}")
+
+    # Korean expansion
+    if result.korean_expansion:
+        print("\n--- Korean Expansion ---")
+        print(f"  Target markets:  {', '.join(result.korean_expansion.target_markets)}")
+        if result.korean_expansion.cjk_url_issues:
+            print(f"  CJK URL issues:  {len(result.korean_expansion.cjk_url_issues)}")
+            for issue in result.korean_expansion.cjk_url_issues[:5]:
+                print(f"    - {issue}")
+        print("  Regional search engines:")
+        for region, engine in result.korean_expansion.regional_search_engines.items():
+            print(f"    [{region}] {engine}")
+        if result.korean_expansion.priority_recommendations:
+            print("  Priority recommendations:")
+            for rec in result.korean_expansion.priority_recommendations[:5]:
+                print(f"    - {rec}")
+
+    # Issues
+    if result.issues:
+        print("\n--- Issues ---")
+        for issue in result.issues:
+            print(f"  - {issue}")
+
+    # Recommendations
+    if result.recommendations:
+        print("\n--- Recommendations ---")
+        for rec in result.recommendations:
+            print(f"  - {rec}")
+
+    print("\n" + "=" * 70)
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def build_parser() -> argparse.ArgumentParser:
+    """Build the argument parser."""
+    parser = argparse.ArgumentParser(
+        description="International SEO Auditor -- analyse multi-language site implementation",
+    )
+    parser.add_argument(
+        "--url",
+        required=True,
+        help="Target URL to audit",
+    )
+    parser.add_argument(
+        "--scope",
+        choices=["structure", "parity", "detection", "redirects", "all"],
+        default="all",
+        help="Audit scope (default: all)",
+    )
+    parser.add_argument(
+        "--korean-expansion",
+        action="store_true",
+        default=False,
+        help="Enable Korean expansion analysis",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        dest="json_output",
+        default=False,
+        help="Output results as JSON",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        default=None,
+        help="Write output to file instead of stdout",
+    )
+    return parser
+
+
+async def async_main(args: argparse.Namespace) -> None:
+    """Async entry point."""
+    auditor = InternationalAuditor(
+        url=args.url,
+        korean_expansion=args.korean_expansion,
+        scope=args.scope,
+    )
+
+    result = await auditor.audit()
+
+    if args.json_output:
+        output = json.dumps(
+            _result_to_dict(result),
+            indent=2,
+            ensure_ascii=False,
+            default=str,
+        )
+    else:
+        # For file output in text mode, capture via print; for stdout, print directly.
+        if args.output:
+            import io
+            buf = io.StringIO()
+            _old_stdout = sys.stdout
+            sys.stdout = buf
+            _print_text_report(result)
+            sys.stdout = _old_stdout
+            output = buf.getvalue()
+        else:
+            _print_text_report(result)
+            return
+
+    if args.output:
+        with open(args.output, "w", encoding="utf-8") as fh:
+            fh.write(output)
+        logger.info("Output written to %s", args.output)
+    else:
+        print(output)
+
+
+def main() -> None:
+    """CLI entry point."""
+    parser = build_parser()
+    args = parser.parse_args()
+    asyncio.run(async_main(args))
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/26-seo-international/code/scripts/requirements.txt b/custom-skills/26-seo-international/code/scripts/requirements.txt
new file mode 100644
index 0000000..c0b8c55
--- /dev/null
+++ b/custom-skills/26-seo-international/code/scripts/requirements.txt
@@ -0,0 +1,10 @@
+# 26-seo-international dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+beautifulsoup4>=4.12.0
+lxml>=5.1.0
+langdetect>=1.0.9
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/26-seo-international/desktop/SKILL.md b/custom-skills/26-seo-international/desktop/SKILL.md
new file mode 100644
index 0000000..dee4255
--- /dev/null
+++ b/custom-skills/26-seo-international/desktop/SKILL.md
@@ -0,0 +1,124 @@
+---
+name: seo-international
+description: |
+  International SEO audit and hreflang validation for multi-language and multi-region websites.
+  Triggers: hreflang, international SEO, multi-language, multi-region, content parity, x-default, ccTLD, 다국어 SEO.
+---
+
+# International SEO Audit
+
+## Purpose
+
+Audit international SEO implementation: hreflang tags, URL structure patterns, content parity across language versions, redirect logic, and Korean expansion strategies. Identify issues preventing proper multi-language indexing.
+
+## Core Capabilities
+
+1. **Hreflang Validation** - Bidirectional links, self-reference, x-default, language code validation
+2. **URL Structure Analysis** - ccTLD vs subdomain vs subdirectory pattern detection
+3. **Content Parity Audit** - Page count comparison, key page availability across languages
+4. **Redirect Logic Audit** - IP-based, Accept-Language redirects, forced redirect detection
+5. **Korean Expansion** - Priority markets (ja, zh, en), CJK URL issues, regional search engines
+
+## MCP Tool Usage
+
+### Ahrefs for Country Metrics
+```
+mcp__ahrefs__site-explorer-metrics-by-country: Country-level traffic distribution
+mcp__ahrefs__site-explorer-organic-keywords: Keywords filtered by country
+```
+
+### Notion for Report Storage
+```
+mcp__notion__notion-create-pages: Save audit report to SEO Audit Log database
+```
+
+### WebSearch for Best Practices
+```
+WebSearch: Research hreflang implementation guides and regional search engine requirements
+```
+
+## Workflow
+
+### 1. Hreflang Validation
+1. Fetch target URL and extract hreflang tags (HTML head, HTTP headers)
+2. If sitemap provided, also extract xhtml:link hreflang from XML sitemap
+3. Validate language codes (ISO 639-1) and region codes (ISO 3166-1)
+4. Check bidirectional links (if A references B, B must reference A)
+5. Verify self-referencing tags on each page
+6. Check x-default tag presence and validity
+7. Detect conflicting hreflang for same language-region
+8. Report all errors with severity levels
+
+### 2. URL Structure Analysis
+1. Crawl known language versions of the site
+2. Classify pattern: ccTLD (example.kr), subdomain (ko.example.com), subdirectory (example.com/ko/)
+3. Check consistency across all language versions
+4. Provide recommendation based on business context
+
+### 3. Content Parity Audit
+1. Discover all language versions from hreflang tags
+2. Count pages per language version
+3. Check availability of key pages (home, about, contact, products/services)
+4. Compare content freshness (last modified dates) across versions
+5. Flag significant gaps in content availability
+
+### 4. Redirect Logic Audit
+1. Test URL with different Accept-Language headers (ko, en, ja, zh)
+2. Check if redirects are forced (no way to override) vs suggested (banner/popup)
+3. Flag forced geo/language redirects as anti-pattern
+4. Recommend proper implementation (suggest, do not force)
+
+### 5. Korean Expansion Analysis (Optional)
+1. Analyze current traffic by country via Ahrefs
+2. Recommend priority target markets for Korean businesses
+3. Check CJK-specific URL encoding issues
+4. Advise on regional search engines (Naver, Baidu, Yahoo Japan)
+
+## Output Format
+
+```markdown
+## 다국어 SEO 감사: [domain]
+
+### Hreflang 검증
+- 검사 페이지 수: [count]
+- 오류: [count] (심각 [count], 경고 [count])
+- 양방향 링크 누락: [list]
+- 자기참조 누락: [list]
+- x-default: [있음/없음]
+
+### URL 구조
+- 패턴: [ccTLD/subdomain/subdirectory]
+- 일관성: [양호/비일관]
+- 권장사항: [recommendation]
+
+### 콘텐츠 동등성
+| 언어 | 페이지 수 | 핵심 페이지 | 최신성 점수 |
+|------|----------|------------|-----------|
+| ko   | 150      | 5/5        | 90        |
+| en   | 120      | 4/5        | 75        |
+
+### 리다이렉트 로직
+- IP 기반 리다이렉트: [있음/없음]
+- 언어 기반 리다이렉트: [있음/없음]
+- 강제 리다이렉트: [있음/없음] (없어야 정상)
+
+### 종합 점수: [score]/100
+
+### 권장 조치사항
+1. [Priority fixes in Korean]
+```
+
+## Notion Output (Required)
+
+All audit reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category (International SEO), Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: INTL-YYYYMMDD-NNN
+
+## Limitations
+
+- Cannot detect server-side IP-based redirects without proxy testing
+- Content language detection requires sufficient text content
+- Large sites (10,000+ pages) require sampling approach
+- Sitemap-based hreflang requires XML sitemap access
diff --git a/custom-skills/26-seo-international/desktop/skill.yaml b/custom-skills/26-seo-international/desktop/skill.yaml
new file mode 100644
index 0000000..6dd0be0
--- /dev/null
+++ b/custom-skills/26-seo-international/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-international
+description: |
+  International SEO audit and hreflang validation. Triggers: hreflang, international SEO, multi-language, multi-region, content parity, x-default, ccTLD.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/26-seo-international/desktop/tools/ahrefs.md b/custom-skills/26-seo-international/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..7becfdf
--- /dev/null
+++ b/custom-skills/26-seo-international/desktop/tools/ahrefs.md
@@ -0,0 +1,43 @@
+# Ahrefs
+
+## Tools Used
+
+### site-explorer-metrics-by-country
+- **Purpose**: Get country-level organic traffic distribution
+- **Usage**: Analyze which countries drive traffic to identify international SEO opportunities
+- **Parameters**: `target` (domain), `country` (optional filter)
+- **Example**:
+  ```
+  mcp__ahrefs__site-explorer-metrics-by-country:
+    target: example.com
+  ```
+
+### site-explorer-organic-keywords
+- **Purpose**: Get organic keyword rankings filtered by country
+- **Usage**: Analyze keyword performance in specific markets
+- **Parameters**: `target` (domain), `country` (ISO country code)
+- **Example**:
+  ```
+  mcp__ahrefs__site-explorer-organic-keywords:
+    target: example.com
+    country: kr
+  ```
+
+## Configuration
+
+- Ahrefs MCP server must be connected in Claude Desktop
+- API access requires active Ahrefs subscription
+
+## Common Patterns
+
+### Country Traffic Analysis
+1. Call `site-explorer-metrics-by-country` to get traffic distribution
+2. Identify top countries by organic traffic share
+3. Compare with hreflang implementation coverage
+4. Flag countries with traffic but no localized version
+
+### Keyword Gap by Market
+1. Call `site-explorer-organic-keywords` with country filter
+2. Compare keyword counts across target markets
+3. Identify markets with low keyword coverage
+4. Recommend content localization priorities
diff --git a/custom-skills/26-seo-international/desktop/tools/notion.md b/custom-skills/26-seo-international/desktop/tools/notion.md
new file mode 100644
index 0000000..fb88957
--- /dev/null
+++ b/custom-skills/26-seo-international/desktop/tools/notion.md
@@ -0,0 +1,51 @@
+# Notion
+
+## Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+## Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title in Korean with date |
+| Site | URL | Audited website URL |
+| Category | Select | "International SEO" |
+| Priority | Select | Based on hreflang error severity |
+| Found Date | Date | Audit date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: INTL-YYYYMMDD-NNN |
+
+## Example: Create Audit Report
+
+```
+mcp__notion__notion-create-pages:
+  pages:
+    - parent_id: "2c8581e5-8a1e-8035-880b-e38cefc2f3ef"
+      parent_type: "database"
+      title: "다국어 SEO 감사 - example.com (2025-01-15)"
+      properties:
+        Site:
+          url: "https://example.com"
+        Category:
+          select:
+            name: "International SEO"
+        Priority:
+          select:
+            name: "High"
+        Found Date:
+          date:
+            start: "2025-01-15"
+        Audit ID:
+          rich_text:
+            - text:
+                content: "INTL-20250115-001"
+```
+
+## Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (hreflang, x-default, ccTLD, subdomain)
+- URLs and code remain unchanged
diff --git a/custom-skills/26-seo-international/desktop/tools/websearch.md b/custom-skills/26-seo-international/desktop/tools/websearch.md
new file mode 100644
index 0000000..9ae4928
--- /dev/null
+++ b/custom-skills/26-seo-international/desktop/tools/websearch.md
@@ -0,0 +1,40 @@
+# WebSearch
+
+## Purpose
+
+Search the web for current international SEO best practices, hreflang implementation guides, and regional search engine requirements.
+
+## Common Search Queries
+
+### Hreflang Best Practices
+```
+WebSearch: "hreflang implementation best practices 2025"
+WebSearch: "hreflang common errors fix"
+WebSearch: "x-default hreflang when to use"
+```
+
+### Regional Search Engines
+```
+WebSearch: "Naver SEO requirements Korean websites"
+WebSearch: "Baidu SEO China market entry"
+WebSearch: "Yahoo Japan SEO vs Google Japan"
+```
+
+### International URL Structure
+```
+WebSearch: "ccTLD vs subdomain vs subdirectory international SEO"
+WebSearch: "Google recommendations international targeting"
+```
+
+### Korean Market Expansion
+```
+WebSearch: "Korean business international SEO Japan market"
+WebSearch: "CJK URL encoding SEO best practices"
+```
+
+## Usage Pattern
+
+1. Search for domain-specific international SEO intelligence
+2. Verify current Google documentation on hreflang
+3. Research regional search engine requirements for target markets
+4. Find competitor international SEO strategies
diff --git a/custom-skills/27-seo-ai-visibility/code/CLAUDE.md b/custom-skills/27-seo-ai-visibility/code/CLAUDE.md
new file mode 100644
index 0000000..96ead6c
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/code/CLAUDE.md
@@ -0,0 +1,147 @@
+# CLAUDE.md
+
+## Overview
+
+AI search visibility and brand radar tool for tracking how a brand appears in AI-generated search answers. Monitors AI answer citations, tracks share of voice in AI search vs competitors, analyzes cited domains and pages, and tracks impressions/mentions history. Uses Ahrefs Brand Radar APIs for comprehensive AI visibility monitoring.
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# AI visibility tracking
+python scripts/ai_visibility_tracker.py --target example.com --json
+
+# AI citation analysis
+python scripts/ai_citation_analyzer.py --target example.com --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `ai_visibility_tracker.py` | Track brand visibility in AI search results | AI impressions, mentions, share of voice, trends |
+| `ai_citation_analyzer.py` | Analyze AI answer citations and source pages | Cited domains, cited pages, AI response analysis |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## AI Visibility Tracker
+
+```bash
+# Current visibility overview
+python scripts/ai_visibility_tracker.py --target example.com --json
+
+# With competitor comparison
+python scripts/ai_visibility_tracker.py --target example.com --competitor comp1.com --competitor comp2.com --json
+
+# Historical trend (impressions/mentions)
+python scripts/ai_visibility_tracker.py --target example.com --history --json
+
+# Share of voice analysis
+python scripts/ai_visibility_tracker.py --target example.com --sov --json
+```
+
+**Capabilities**:
+- AI impressions overview (how often brand appears in AI answers)
+- AI mentions overview (brand mention frequency across AI engines)
+- Share of Voice in AI search vs competitors
+- Impressions history over time (trend tracking)
+- Mentions history over time
+- SOV history and trend analysis
+- Competitor AI visibility comparison
+
+## AI Citation Analyzer
+
+```bash
+# Analyze AI citations for brand
+python scripts/ai_citation_analyzer.py --target example.com --json
+
+# Cited domains analysis
+python scripts/ai_citation_analyzer.py --target example.com --cited-domains --json
+
+# Cited pages analysis
+python scripts/ai_citation_analyzer.py --target example.com --cited-pages --json
+
+# AI response content analysis
+python scripts/ai_citation_analyzer.py --target example.com --responses --json
+```
+
+**Capabilities**:
+- AI response analysis (how the brand appears in AI-generated answers)
+- Cited domains analysis (which source domains AI engines reference)
+- Cited pages analysis (which specific URLs get cited)
+- Citation sentiment and context analysis
+- Citation frequency ranking
+- Competitor citation comparison
+- Recommendation generation for improving AI visibility
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `brand-radar-ai-responses` | Get AI-generated responses mentioning brand |
+| `brand-radar-cited-domains` | Domains cited in AI answers |
+| `brand-radar-cited-pages` | Specific pages cited in AI answers |
+| `brand-radar-impressions-history` | Brand impression trend over time |
+| `brand-radar-impressions-overview` | Current impression metrics |
+| `brand-radar-mentions-history` | Brand mention trend over time |
+| `brand-radar-mentions-overview` | Current mention metrics |
+| `brand-radar-sov-history` | Share of voice trend |
+| `brand-radar-sov-overview` | Current share of voice |
+
+## Output Format
+
+```json
+{
+  "target": "example.com",
+  "impressions": {
+    "total": 15000,
+    "trend": "increasing",
+    "change_pct": 12.5
+  },
+  "mentions": {
+    "total": 850,
+    "trend": "stable",
+    "change_pct": 2.1
+  },
+  "share_of_voice": {
+    "brand_sov": 18.5,
+    "competitors": [
+      {"domain": "comp1.com", "sov": 25.3},
+      {"domain": "comp2.com", "sov": 15.8}
+    ]
+  },
+  "cited_domains": [...],
+  "cited_pages": [...],
+  "ai_responses_sample": [...],
+  "recommendations": [...],
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Tracked website URL |
+| Category | Select | AI Search Visibility |
+| Priority | Select | Based on SOV trend |
+| Found Date | Date | Report date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: AI-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., AI Search, Share of Voice, Brand Radar)
+- URLs and code remain unchanged
diff --git a/custom-skills/27-seo-ai-visibility/code/scripts/ai_citation_analyzer.py b/custom-skills/27-seo-ai-visibility/code/scripts/ai_citation_analyzer.py
new file mode 100644
index 0000000..08e88e8
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/code/scripts/ai_citation_analyzer.py
@@ -0,0 +1,611 @@
+"""
+AI Citation Analyzer - Brand Radar Citation Analysis
+=====================================================
+Purpose: Analyze how a brand is cited in AI-generated search answers,
+         including cited domains, cited pages, and AI response content.
+Python: 3.10+
+
+Usage:
+    python ai_citation_analyzer.py --target example.com --json
+    python ai_citation_analyzer.py --target example.com --cited-domains --json
+    python ai_citation_analyzer.py --target example.com --cited-pages --json
+    python ai_citation_analyzer.py --target example.com --responses --json
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import subprocess
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from pathlib import Path
+from typing import Any
+
+# Add parent to path for base_client import
+sys.path.insert(0, str(Path(__file__).parent))
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class AiResponse:
+    """An AI-generated response that mentions the brand."""
+    query: str = ""
+    response_text: str = ""
+    brand_mentioned: bool = False
+    sentiment: str = "neutral"  # positive, neutral, negative
+    source_engine: str = ""
+    date: str = ""
+    url: str = ""
+
+
+@dataclass
+class CitedDomain:
+    """A domain cited in AI-generated answers."""
+    domain: str = ""
+    citation_count: int = 0
+    topics: list[str] = field(default_factory=list)
+    share_pct: float = 0.0
+
+
+@dataclass
+class CitedPage:
+    """A specific page cited in AI-generated answers."""
+    url: str = ""
+    title: str = ""
+    citation_count: int = 0
+    context: str = ""
+    topics: list[str] = field(default_factory=list)
+
+
+@dataclass
+class CitationAnalysisResult:
+    """Complete citation analysis result."""
+    target: str = ""
+    ai_responses: list[AiResponse] = field(default_factory=list)
+    cited_domains: list[CitedDomain] = field(default_factory=list)
+    cited_pages: list[CitedPage] = field(default_factory=list)
+    sentiment_summary: dict = field(default_factory=dict)
+    citation_ranking: list[dict] = field(default_factory=list)
+    competitor_citations: list[dict] = field(default_factory=list)
+    recommendations: list[str] = field(default_factory=list)
+    timestamp: str = field(default_factory=lambda: datetime.now().isoformat())
+
+    def to_dict(self) -> dict:
+        """Convert result to dictionary."""
+        return {
+            "target": self.target,
+            "ai_responses": [asdict(r) for r in self.ai_responses],
+            "cited_domains": [asdict(d) for d in self.cited_domains],
+            "cited_pages": [asdict(p) for p in self.cited_pages],
+            "sentiment_summary": self.sentiment_summary,
+            "citation_ranking": self.citation_ranking,
+            "competitor_citations": self.competitor_citations,
+            "recommendations": self.recommendations,
+            "timestamp": self.timestamp,
+        }
+
+
+# ---------------------------------------------------------------------------
+# MCP tool caller helper
+# ---------------------------------------------------------------------------
+
+def call_mcp_tool(tool_name: str, params: dict) -> dict:
+    """
+    Call an Ahrefs MCP tool and return the parsed JSON response.
+
+    In Claude Desktop / Claude Code environments the MCP tools are invoked
+    directly by the AI agent. This helper exists so that the script can also
+    be executed standalone via subprocess for testing purposes.
+    """
+    logger.info(f"Calling MCP tool: {tool_name} with params: {params}")
+    try:
+        cmd = ["claude", "mcp", "call", "ahrefs", tool_name, json.dumps(params)]
+        result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
+        if result.returncode == 0 and result.stdout.strip():
+            return json.loads(result.stdout.strip())
+        logger.warning(f"MCP tool {tool_name} returned non-zero or empty: {result.stderr}")
+        return {}
+    except (subprocess.TimeoutExpired, json.JSONDecodeError, FileNotFoundError) as exc:
+        logger.warning(f"MCP call failed ({exc}). Returning empty dict.")
+        return {}
+
+
+# ---------------------------------------------------------------------------
+# AI Citation Analyzer
+# ---------------------------------------------------------------------------
+
+class AiCitationAnalyzer(BaseAsyncClient):
+    """Analyze AI answer citations and source pages for a brand."""
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.logger = logging.getLogger(self.__class__.__name__)
+
+    # ---- AI Responses ----
+
+    async def get_ai_responses(self, target: str) -> list[AiResponse]:
+        """Fetch AI-generated responses mentioning the brand via brand-radar-ai-responses."""
+        self.logger.info(f"Fetching AI responses for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-ai-responses",
+            {"target": target},
+        )
+        responses: list[AiResponse] = []
+        if not data:
+            return responses
+
+        items = data if isinstance(data, list) else data.get("responses", data.get("data", []))
+        for item in items:
+            if isinstance(item, dict):
+                responses.append(AiResponse(
+                    query=item.get("query", item.get("keyword", "")),
+                    response_text=item.get("response_text", item.get("answer", item.get("text", ""))),
+                    brand_mentioned=item.get("brand_mentioned", True),
+                    sentiment=item.get("sentiment", "neutral"),
+                    source_engine=item.get("source_engine", item.get("engine", "")),
+                    date=item.get("date", ""),
+                    url=item.get("url", ""),
+                ))
+        return responses
+
+    # ---- Cited Domains ----
+
+    async def get_cited_domains(self, target: str) -> list[CitedDomain]:
+        """Fetch domains cited in AI answers via brand-radar-cited-domains."""
+        self.logger.info(f"Fetching cited domains for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-cited-domains",
+            {"target": target},
+        )
+        domains: list[CitedDomain] = []
+        if not data:
+            return domains
+
+        items = data if isinstance(data, list) else data.get("domains", data.get("data", []))
+        for item in items:
+            if isinstance(item, dict):
+                domains.append(CitedDomain(
+                    domain=item.get("domain", ""),
+                    citation_count=item.get("citation_count", item.get("citations", item.get("count", 0))),
+                    topics=item.get("topics", []),
+                    share_pct=item.get("share_pct", item.get("share", 0.0)),
+                ))
+        return domains
+
+    # ---- Cited Pages ----
+
+    async def get_cited_pages(self, target: str) -> list[CitedPage]:
+        """Fetch specific pages cited in AI answers via brand-radar-cited-pages."""
+        self.logger.info(f"Fetching cited pages for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-cited-pages",
+            {"target": target},
+        )
+        pages: list[CitedPage] = []
+        if not data:
+            return pages
+
+        items = data if isinstance(data, list) else data.get("pages", data.get("data", []))
+        for item in items:
+            if isinstance(item, dict):
+                pages.append(CitedPage(
+                    url=item.get("url", ""),
+                    title=item.get("title", ""),
+                    citation_count=item.get("citation_count", item.get("citations", item.get("count", 0))),
+                    context=item.get("context", item.get("snippet", "")),
+                    topics=item.get("topics", []),
+                ))
+        return pages
+
+    # ---- Sentiment Analysis ----
+
+    @staticmethod
+    def analyze_response_sentiment(responses: list[AiResponse]) -> dict:
+        """
+        Analyze the sentiment distribution of AI responses.
+
+        Returns a summary with counts and percentages for each sentiment category.
+        """
+        if not responses:
+            return {
+                "total": 0,
+                "positive": 0,
+                "neutral": 0,
+                "negative": 0,
+                "positive_pct": 0.0,
+                "neutral_pct": 0.0,
+                "negative_pct": 0.0,
+                "overall_sentiment": "unknown",
+            }
+
+        total = len(responses)
+        positive = sum(1 for r in responses if r.sentiment == "positive")
+        neutral = sum(1 for r in responses if r.sentiment == "neutral")
+        negative = sum(1 for r in responses if r.sentiment == "negative")
+
+        positive_pct = round((positive / total) * 100, 1)
+        neutral_pct = round((neutral / total) * 100, 1)
+        negative_pct = round((negative / total) * 100, 1)
+
+        # Determine overall sentiment
+        if positive_pct >= 60:
+            overall = "positive"
+        elif negative_pct >= 40:
+            overall = "negative"
+        elif positive_pct > negative_pct:
+            overall = "leaning_positive"
+        elif negative_pct > positive_pct:
+            overall = "leaning_negative"
+        else:
+            overall = "neutral"
+
+        return {
+            "total": total,
+            "positive": positive,
+            "neutral": neutral,
+            "negative": negative,
+            "positive_pct": positive_pct,
+            "neutral_pct": neutral_pct,
+            "negative_pct": negative_pct,
+            "overall_sentiment": overall,
+        }
+
+    # ---- Citation Ranking ----
+
+    @staticmethod
+    def rank_citations(items: list[CitedDomain] | list[CitedPage]) -> list[dict]:
+        """Rank cited domains or pages by citation frequency."""
+        if not items:
+            return []
+
+        ranked = sorted(items, key=lambda x: x.citation_count, reverse=True)
+        total_citations = sum(item.citation_count for item in ranked)
+
+        result = []
+        for rank, item in enumerate(ranked, 1):
+            entry = asdict(item)
+            entry["rank"] = rank
+            entry["share_of_citations"] = (
+                round((item.citation_count / total_citations) * 100, 1)
+                if total_citations > 0
+                else 0.0
+            )
+            result.append(entry)
+
+        return result
+
+    # ---- Competitor Citation Comparison ----
+
+    async def compare_competitor_citations(
+        self, target: str, competitors: list[str]
+    ) -> list[dict]:
+        """Compare citation profiles between target and competitors."""
+        self.logger.info(f"Comparing citations for {target} vs {competitors}")
+        results = []
+
+        all_domains = [target] + competitors
+        for domain in all_domains:
+            cited_domains = await self.get_cited_domains(domain)
+            cited_pages = await self.get_cited_pages(domain)
+
+            total_domain_citations = sum(d.citation_count for d in cited_domains)
+            total_page_citations = sum(p.citation_count for p in cited_pages)
+            unique_domains = len(cited_domains)
+            unique_pages = len(cited_pages)
+
+            results.append({
+                "domain": domain,
+                "is_target": domain == target,
+                "total_domain_citations": total_domain_citations,
+                "total_page_citations": total_page_citations,
+                "unique_cited_domains": unique_domains,
+                "unique_cited_pages": unique_pages,
+                "top_cited_domain": cited_domains[0].domain if cited_domains else "",
+                "top_cited_page": cited_pages[0].url if cited_pages else "",
+            })
+
+        # Sort by total page citations descending
+        results.sort(key=lambda x: x["total_page_citations"], reverse=True)
+        return results
+
+    # ---- Recommendations ----
+
+    @staticmethod
+    def generate_recommendations(result: CitationAnalysisResult) -> list[str]:
+        """Generate actionable recommendations for improving AI citations."""
+        recs: list[str] = []
+
+        # Based on citation count
+        total_page_citations = sum(p.citation_count for p in result.cited_pages)
+        if total_page_citations == 0:
+            recs.append(
+                "AI 검색 엔진에서 인용된 페이지가 없습니다. "
+                "고품질 원본 콘텐츠(연구 데이터, 종합 가이드, 전문가 인사이트)를 "
+                "발행하여 AI 엔진의 인용 대상이 되도록 하세요."
+            )
+        elif total_page_citations < 10:
+            recs.append(
+                f"인용된 페이지 수가 {total_page_citations}건으로 적습니다. "
+                "FAQ, How-to, 비교 분석 등 AI가 참조하기 쉬운 "
+                "구조화된 콘텐츠를 추가하세요."
+            )
+
+        # Based on domain diversity
+        if result.cited_domains:
+            target_domains = [d for d in result.cited_domains if d.domain == result.target]
+            if not target_domains:
+                recs.append(
+                    "타깃 도메인이 AI 인용 도메인 목록에 포함되지 않았습니다. "
+                    "도메인 권위(Domain Authority) 향상과 "
+                    "Schema Markup(JSON-LD) 적용을 우선 추진하세요."
+                )
+
+        # Based on sentiment
+        sentiment = result.sentiment_summary
+        if sentiment.get("negative_pct", 0) > 30:
+            recs.append(
+                f"AI 응답 중 부정적 언급 비율이 {sentiment['negative_pct']}%입니다. "
+                "브랜드 평판 관리와 긍정적 콘텐츠 확대가 필요합니다. "
+                "고객 리뷰, 성공 사례, 수상 내역 등을 강화하세요."
+            )
+        elif sentiment.get("overall_sentiment") == "positive":
+            recs.append(
+                "AI 응답에서 브랜드 언급이 전반적으로 긍정적입니다. "
+                "이 긍정적 이미지를 활용하여 더 많은 키워드에서 "
+                "AI 인용을 확대하세요."
+            )
+
+        # Content strategy recommendations
+        if result.cited_pages:
+            top_pages = sorted(result.cited_pages, key=lambda p: p.citation_count, reverse=True)[:3]
+            top_topics = set()
+            for page in top_pages:
+                top_topics.update(page.topics)
+            if top_topics:
+                topics_str = ", ".join(list(top_topics)[:5])
+                recs.append(
+                    f"가장 많이 인용되는 주제는 [{topics_str}]입니다. "
+                    "이 주제들에 대한 심층 콘텐츠를 추가 제작하세요."
+                )
+
+        # E-E-A-T and structured data
+        recs.append(
+            "AI 인용률 향상을 위한 핵심 전략: "
+            "(1) E-E-A-T 시그널 강화 - 저자 프로필, 전문가 인용, 실제 경험 콘텐츠, "
+            "(2) 구조화된 데이터 적용 - FAQ, HowTo, Article Schema, "
+            "(3) 콘텐츠 정확성 및 최신성 유지, "
+            "(4) 원본 데이터와 독자적 연구 결과 발행."
+        )
+
+        # Competitor-based recommendations
+        if result.competitor_citations:
+            leader = result.competitor_citations[0]
+            if not leader.get("is_target", False):
+                recs.append(
+                    f"인용 리더는 {leader['domain']}입니다 "
+                    f"(페이지 인용 {leader['total_page_citations']}건). "
+                    "해당 경쟁사의 인용된 페이지를 분석하여 "
+                    "콘텐츠 갭을 파악하세요."
+                )
+
+        return recs
+
+    # ---- Main Orchestrator ----
+
+    async def analyze(
+        self,
+        target: str,
+        competitors: list[str] | None = None,
+        include_responses: bool = True,
+        include_cited_domains: bool = True,
+        include_cited_pages: bool = True,
+    ) -> CitationAnalysisResult:
+        """
+        Orchestrate full citation analysis.
+
+        Args:
+            target: Domain to analyze
+            competitors: Optional competitor domains
+            include_responses: Fetch AI response data
+            include_cited_domains: Fetch cited domains
+            include_cited_pages: Fetch cited pages
+        """
+        self.logger.info(f"Starting AI citation analysis for {target}")
+        result = CitationAnalysisResult(target=target)
+
+        # AI responses
+        if include_responses:
+            result.ai_responses = await self.get_ai_responses(target)
+            result.sentiment_summary = self.analyze_response_sentiment(result.ai_responses)
+
+        # Cited domains
+        if include_cited_domains:
+            result.cited_domains = await self.get_cited_domains(target)
+            if result.cited_domains:
+                result.citation_ranking = self.rank_citations(result.cited_domains)
+
+        # Cited pages
+        if include_cited_pages:
+            result.cited_pages = await self.get_cited_pages(target)
+
+        # Competitor comparison
+        if competitors:
+            result.competitor_citations = await self.compare_competitor_citations(
+                target, competitors
+            )
+
+        # Recommendations
+        result.recommendations = self.generate_recommendations(result)
+
+        self.print_stats()
+        return result
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def build_parser() -> argparse.ArgumentParser:
+    """Build argument parser for CLI usage."""
+    parser = argparse.ArgumentParser(
+        description="AI Citation Analyzer - Analyze AI answer citations and source pages",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  %(prog)s --target example.com --json
+  %(prog)s --target example.com --cited-domains --json
+  %(prog)s --target example.com --cited-pages --json
+  %(prog)s --target example.com --responses --competitor comp1.com --json
+  %(prog)s --target example.com --output citations.json
+        """,
+    )
+    parser.add_argument(
+        "--target", required=True,
+        help="Target domain to analyze (e.g., example.com)",
+    )
+    parser.add_argument(
+        "--competitor", action="append", default=[],
+        help="Competitor domain (repeatable). e.g., --competitor a.com --competitor b.com",
+    )
+    parser.add_argument(
+        "--cited-domains", action="store_true",
+        help="Include cited domains analysis",
+    )
+    parser.add_argument(
+        "--cited-pages", action="store_true",
+        help="Include cited pages analysis",
+    )
+    parser.add_argument(
+        "--responses", action="store_true",
+        help="Include AI response content analysis",
+    )
+    parser.add_argument(
+        "--json", action="store_true",
+        help="Output result as JSON to stdout",
+    )
+    parser.add_argument(
+        "--output", type=str, default=None,
+        help="Save JSON output to file path",
+    )
+    return parser
+
+
+def print_summary(result: CitationAnalysisResult) -> None:
+    """Print a human-readable summary of citation analysis."""
+    print("\n" + "=" * 60)
+    print(f"  AI Citation Analysis: {result.target}")
+    print("=" * 60)
+
+    # AI Responses
+    if result.ai_responses:
+        print(f"\n  AI Responses: {len(result.ai_responses)}")
+        for resp in result.ai_responses[:5]:
+            engine_tag = f" [{resp.source_engine}]" if resp.source_engine else ""
+            sentiment_tag = f" ({resp.sentiment})"
+            print(f"    - Q: {resp.query[:60]}{engine_tag}{sentiment_tag}")
+        if len(result.ai_responses) > 5:
+            print(f"    ... and {len(result.ai_responses) - 5} more")
+
+    # Sentiment Summary
+    if result.sentiment_summary:
+        s = result.sentiment_summary
+        print(f"\n  Sentiment: {s.get('overall_sentiment', 'unknown')}")
+        print(f"    Positive: {s.get('positive', 0)} ({s.get('positive_pct', 0):.1f}%)")
+        print(f"    Neutral:  {s.get('neutral', 0)} ({s.get('neutral_pct', 0):.1f}%)")
+        print(f"    Negative: {s.get('negative', 0)} ({s.get('negative_pct', 0):.1f}%)")
+
+    # Cited Domains
+    if result.cited_domains:
+        print(f"\n  Cited Domains: {len(result.cited_domains)}")
+        for domain in result.cited_domains[:10]:
+            topics_str = ", ".join(domain.topics[:3]) if domain.topics else ""
+            print(f"    {domain.domain}: {domain.citation_count} citations"
+                  f"{f' [{topics_str}]' if topics_str else ''}")
+        if len(result.cited_domains) > 10:
+            print(f"    ... and {len(result.cited_domains) - 10} more")
+
+    # Cited Pages
+    if result.cited_pages:
+        print(f"\n  Cited Pages: {len(result.cited_pages)}")
+        for page in result.cited_pages[:10]:
+            title = page.title[:50] if page.title else page.url[:50]
+            print(f"    {title}: {page.citation_count} citations")
+        if len(result.cited_pages) > 10:
+            print(f"    ... and {len(result.cited_pages) - 10} more")
+
+    # Competitor Comparison
+    if result.competitor_citations:
+        print("\n  Competitor Citation Comparison:")
+        for comp in result.competitor_citations:
+            marker = " <-- target" if comp.get("is_target") else ""
+            print(f"    {comp['domain']}: "
+                  f"domains={comp['unique_cited_domains']}, "
+                  f"pages={comp['unique_cited_pages']}, "
+                  f"page_citations={comp['total_page_citations']}{marker}")
+
+    # Recommendations
+    if result.recommendations:
+        print("\n  Recommendations:")
+        for i, rec in enumerate(result.recommendations, 1):
+            print(f"    {i}. {rec}")
+
+    print("\n" + "=" * 60)
+    print(f"  Generated: {result.timestamp}")
+    print("=" * 60 + "\n")
+
+
+async def main() -> None:
+    """CLI entry point."""
+    parser = build_parser()
+    args = parser.parse_args()
+
+    # Determine which sections to include
+    # If no specific flags, include everything
+    any_specific = args.cited_domains or args.cited_pages or args.responses
+    include_responses = args.responses or not any_specific
+    include_cited_domains = args.cited_domains or not any_specific
+    include_cited_pages = args.cited_pages or not any_specific
+
+    analyzer = AiCitationAnalyzer(
+        max_concurrent=5,
+        requests_per_second=2.0,
+    )
+
+    result = await analyzer.analyze(
+        target=args.target,
+        competitors=args.competitor if args.competitor else None,
+        include_responses=include_responses,
+        include_cited_domains=include_cited_domains,
+        include_cited_pages=include_cited_pages,
+    )
+
+    # Output
+    if args.json or args.output:
+        output_data = result.to_dict()
+        json_str = json.dumps(output_data, ensure_ascii=False, indent=2)
+
+        if args.json:
+            print(json_str)
+
+        if args.output:
+            output_path = Path(args.output)
+            output_path.parent.mkdir(parents=True, exist_ok=True)
+            output_path.write_text(json_str, encoding="utf-8")
+            logger.info(f"Report saved to {args.output}")
+    else:
+        print_summary(result)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/27-seo-ai-visibility/code/scripts/ai_visibility_tracker.py b/custom-skills/27-seo-ai-visibility/code/scripts/ai_visibility_tracker.py
new file mode 100644
index 0000000..10dc023
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/code/scripts/ai_visibility_tracker.py
@@ -0,0 +1,594 @@
+"""
+AI Visibility Tracker - Brand Radar Monitoring
+================================================
+Purpose: Track brand visibility in AI-generated search answers
+         using Ahrefs Brand Radar APIs.
+Python: 3.10+
+
+Usage:
+    python ai_visibility_tracker.py --target example.com --json
+    python ai_visibility_tracker.py --target example.com --competitor comp1.com --json
+    python ai_visibility_tracker.py --target example.com --history --json
+    python ai_visibility_tracker.py --target example.com --sov --json
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import subprocess
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from pathlib import Path
+from typing import Any
+
+# Add parent to path for base_client import
+sys.path.insert(0, str(Path(__file__).parent))
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class ImpressionMetrics:
+    """AI search impression metrics for a brand."""
+    total: int = 0
+    trend: str = "stable"  # increasing, decreasing, stable
+    change_pct: float = 0.0
+    period: str = ""
+    breakdown: dict = field(default_factory=dict)
+
+
+@dataclass
+class MentionMetrics:
+    """AI search mention metrics for a brand."""
+    total: int = 0
+    trend: str = "stable"
+    change_pct: float = 0.0
+    period: str = ""
+    breakdown: dict = field(default_factory=dict)
+
+
+@dataclass
+class SovMetric:
+    """Share of Voice metric for a single domain."""
+    domain: str = ""
+    sov_pct: float = 0.0
+    change_pct: float = 0.0
+
+
+@dataclass
+class HistoryPoint:
+    """Single data point in a time series."""
+    date: str = ""
+    value: float = 0.0
+
+
+@dataclass
+class CompetitorVisibility:
+    """Aggregated AI visibility metrics for a competitor domain."""
+    domain: str = ""
+    impressions: int = 0
+    mentions: int = 0
+    sov: float = 0.0
+
+
+@dataclass
+class AiVisibilityResult:
+    """Complete AI visibility tracking result."""
+    target: str = ""
+    impressions: ImpressionMetrics = field(default_factory=ImpressionMetrics)
+    mentions: MentionMetrics = field(default_factory=MentionMetrics)
+    share_of_voice: dict = field(default_factory=dict)
+    impressions_history: list[HistoryPoint] = field(default_factory=list)
+    mentions_history: list[HistoryPoint] = field(default_factory=list)
+    sov_history: list[HistoryPoint] = field(default_factory=list)
+    competitors: list[CompetitorVisibility] = field(default_factory=list)
+    recommendations: list[str] = field(default_factory=list)
+    timestamp: str = field(default_factory=lambda: datetime.now().isoformat())
+
+    def to_dict(self) -> dict:
+        """Convert result to dictionary."""
+        return {
+            "target": self.target,
+            "impressions": asdict(self.impressions),
+            "mentions": asdict(self.mentions),
+            "share_of_voice": self.share_of_voice,
+            "impressions_history": [asdict(h) for h in self.impressions_history],
+            "mentions_history": [asdict(h) for h in self.mentions_history],
+            "sov_history": [asdict(h) for h in self.sov_history],
+            "competitors": [asdict(c) for c in self.competitors],
+            "recommendations": self.recommendations,
+            "timestamp": self.timestamp,
+        }
+
+
+# ---------------------------------------------------------------------------
+# MCP tool caller helper
+# ---------------------------------------------------------------------------
+
+def call_mcp_tool(tool_name: str, params: dict) -> dict:
+    """
+    Call an Ahrefs MCP tool and return the parsed JSON response.
+
+    In Claude Desktop / Claude Code environments the MCP tools are invoked
+    directly by the AI agent. This helper exists so that the script can also
+    be executed standalone via subprocess for testing purposes.
+    """
+    logger.info(f"Calling MCP tool: {tool_name} with params: {params}")
+    try:
+        cmd = ["claude", "mcp", "call", "ahrefs", tool_name, json.dumps(params)]
+        result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
+        if result.returncode == 0 and result.stdout.strip():
+            return json.loads(result.stdout.strip())
+        logger.warning(f"MCP tool {tool_name} returned non-zero or empty: {result.stderr}")
+        return {}
+    except (subprocess.TimeoutExpired, json.JSONDecodeError, FileNotFoundError) as exc:
+        logger.warning(f"MCP call failed ({exc}). Returning empty dict.")
+        return {}
+
+
+# ---------------------------------------------------------------------------
+# AI Visibility Tracker
+# ---------------------------------------------------------------------------
+
+class AiVisibilityTracker(BaseAsyncClient):
+    """Track brand visibility across AI-generated search results."""
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.logger = logging.getLogger(self.__class__.__name__)
+
+    # ---- Impressions ----
+
+    async def get_impressions_overview(self, target: str) -> ImpressionMetrics:
+        """Fetch current AI impression metrics via brand-radar-impressions-overview."""
+        self.logger.info(f"Fetching impressions overview for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-impressions-overview",
+            {"target": target},
+        )
+        metrics = ImpressionMetrics()
+        if not data:
+            return metrics
+
+        metrics.total = data.get("total_impressions", data.get("impressions", 0))
+        metrics.change_pct = data.get("change_pct", data.get("change", 0.0))
+        metrics.period = data.get("period", "")
+        metrics.breakdown = data.get("breakdown", {})
+
+        if metrics.change_pct > 5:
+            metrics.trend = "increasing"
+        elif metrics.change_pct < -5:
+            metrics.trend = "decreasing"
+        else:
+            metrics.trend = "stable"
+
+        return metrics
+
+    # ---- Mentions ----
+
+    async def get_mentions_overview(self, target: str) -> MentionMetrics:
+        """Fetch current AI mention metrics via brand-radar-mentions-overview."""
+        self.logger.info(f"Fetching mentions overview for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-mentions-overview",
+            {"target": target},
+        )
+        metrics = MentionMetrics()
+        if not data:
+            return metrics
+
+        metrics.total = data.get("total_mentions", data.get("mentions", 0))
+        metrics.change_pct = data.get("change_pct", data.get("change", 0.0))
+        metrics.period = data.get("period", "")
+        metrics.breakdown = data.get("breakdown", {})
+
+        if metrics.change_pct > 5:
+            metrics.trend = "increasing"
+        elif metrics.change_pct < -5:
+            metrics.trend = "decreasing"
+        else:
+            metrics.trend = "stable"
+
+        return metrics
+
+    # ---- Share of Voice ----
+
+    async def get_sov_overview(self, target: str) -> dict:
+        """Fetch Share of Voice overview via brand-radar-sov-overview."""
+        self.logger.info(f"Fetching SOV overview for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-sov-overview",
+            {"target": target},
+        )
+        if not data:
+            return {"brand_sov": 0.0, "competitors": []}
+
+        brand_sov = data.get("sov", data.get("share_of_voice", 0.0))
+        competitors_raw = data.get("competitors", [])
+        competitors = []
+        for comp in competitors_raw:
+            competitors.append(SovMetric(
+                domain=comp.get("domain", ""),
+                sov_pct=comp.get("sov", comp.get("share_of_voice", 0.0)),
+                change_pct=comp.get("change_pct", 0.0),
+            ))
+
+        return {
+            "brand_sov": brand_sov,
+            "competitors": [asdict(c) for c in competitors],
+        }
+
+    # ---- History ----
+
+    async def get_impressions_history(self, target: str) -> list[HistoryPoint]:
+        """Fetch impressions history via brand-radar-impressions-history."""
+        self.logger.info(f"Fetching impressions history for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-impressions-history",
+            {"target": target},
+        )
+        return self._parse_history(data)
+
+    async def get_mentions_history(self, target: str) -> list[HistoryPoint]:
+        """Fetch mentions history via brand-radar-mentions-history."""
+        self.logger.info(f"Fetching mentions history for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-mentions-history",
+            {"target": target},
+        )
+        return self._parse_history(data)
+
+    async def get_sov_history(self, target: str) -> list[HistoryPoint]:
+        """Fetch SOV history via brand-radar-sov-history."""
+        self.logger.info(f"Fetching SOV history for {target}")
+        data = await asyncio.to_thread(
+            call_mcp_tool,
+            "brand-radar-sov-history",
+            {"target": target},
+        )
+        return self._parse_history(data)
+
+    def _parse_history(self, data: dict | list) -> list[HistoryPoint]:
+        """Parse history data from MCP response into HistoryPoint list."""
+        points: list[HistoryPoint] = []
+        if not data:
+            return points
+
+        items = data if isinstance(data, list) else data.get("history", data.get("data", []))
+        for item in items:
+            if isinstance(item, dict):
+                points.append(HistoryPoint(
+                    date=item.get("date", item.get("period", "")),
+                    value=item.get("value", item.get("impressions", item.get("mentions", item.get("sov", 0.0)))),
+                ))
+        return points
+
+    # ---- Competitor Comparison ----
+
+    async def compare_competitors(
+        self, target: str, competitors: list[str]
+    ) -> list[CompetitorVisibility]:
+        """Aggregate AI visibility metrics for target and competitors."""
+        self.logger.info(f"Comparing competitors: {competitors}")
+        results: list[CompetitorVisibility] = []
+
+        all_domains = [target] + competitors
+        for domain in all_domains:
+            imp = await self.get_impressions_overview(domain)
+            men = await self.get_mentions_overview(domain)
+            sov_data = await self.get_sov_overview(domain)
+
+            results.append(CompetitorVisibility(
+                domain=domain,
+                impressions=imp.total,
+                mentions=men.total,
+                sov=sov_data.get("brand_sov", 0.0),
+            ))
+
+        # Sort by SOV descending
+        results.sort(key=lambda x: x.sov, reverse=True)
+        return results
+
+    # ---- Trend Calculation ----
+
+    @staticmethod
+    def calculate_trends(history: list[HistoryPoint]) -> dict:
+        """Determine trend direction and statistics from history data."""
+        if not history or len(history) < 2:
+            return {
+                "direction": "insufficient_data",
+                "avg_value": 0.0,
+                "min_value": 0.0,
+                "max_value": 0.0,
+                "change_pct": 0.0,
+                "data_points": len(history) if history else 0,
+            }
+
+        values = [h.value for h in history]
+        first_value = values[0]
+        last_value = values[-1]
+        avg_value = sum(values) / len(values)
+        min_value = min(values)
+        max_value = max(values)
+
+        if first_value > 0:
+            change_pct = ((last_value - first_value) / first_value) * 100
+        else:
+            change_pct = 0.0
+
+        if change_pct > 10:
+            direction = "strongly_increasing"
+        elif change_pct > 3:
+            direction = "increasing"
+        elif change_pct < -10:
+            direction = "strongly_decreasing"
+        elif change_pct < -3:
+            direction = "decreasing"
+        else:
+            direction = "stable"
+
+        return {
+            "direction": direction,
+            "avg_value": round(avg_value, 2),
+            "min_value": round(min_value, 2),
+            "max_value": round(max_value, 2),
+            "change_pct": round(change_pct, 2),
+            "data_points": len(values),
+        }
+
+    # ---- Recommendations ----
+
+    @staticmethod
+    def generate_recommendations(result: AiVisibilityResult) -> list[str]:
+        """Generate actionable recommendations for improving AI visibility."""
+        recs: list[str] = []
+
+        # Impression-based recommendations
+        if result.impressions.total == 0:
+            recs.append(
+                "AI 검색에서 브랜드 노출이 감지되지 않았습니다. "
+                "E-E-A-T 시그널(경험, 전문성, 권위성, 신뢰성)을 강화하여 "
+                "AI 엔진이 콘텐츠를 참조할 수 있도록 하세요."
+            )
+        elif result.impressions.trend == "decreasing":
+            recs.append(
+                "AI 검색 노출이 감소 추세입니다. 최신 콘텐츠 업데이트 및 "
+                "구조화된 데이터(Schema Markup) 추가를 검토하세요."
+            )
+        elif result.impressions.trend == "increasing":
+            recs.append(
+                "AI 검색 노출이 증가 추세입니다. 현재 콘텐츠 전략을 "
+                "유지하면서 추가 키워드 확장을 고려하세요."
+            )
+
+        # Mention-based recommendations
+        if result.mentions.total == 0:
+            recs.append(
+                "AI 응답에서 브랜드 언급이 없습니다. "
+                "브랜드명이 포함된 고품질 콘텐츠를 제작하고, "
+                "외부 사이트에서의 브랜드 언급(Citations)을 늘리세요."
+            )
+        elif result.mentions.trend == "decreasing":
+            recs.append(
+                "AI 응답 내 브랜드 언급이 줄어들고 있습니다. "
+                "콘텐츠 신선도(Freshness)와 업계 권위 신호를 점검하세요."
+            )
+
+        # SOV recommendations
+        sov_value = result.share_of_voice.get("brand_sov", 0.0)
+        if sov_value < 10:
+            recs.append(
+                f"AI 검색 Share of Voice가 {sov_value}%로 낮습니다. "
+                "핵심 키워드에 대한 종합 가이드, FAQ 콘텐츠, "
+                "원본 데이터/연구 자료를 발행하여 인용 가능성을 높이세요."
+            )
+        elif sov_value < 25:
+            recs.append(
+                f"AI 검색 Share of Voice가 {sov_value}%입니다. "
+                "경쟁사 대비 차별화된 전문 콘텐츠와 "
+                "독점 데이터 기반 인사이트를 강화하세요."
+            )
+
+        # Competitor-based recommendations
+        if result.competitors:
+            top_competitor = result.competitors[0]
+            if top_competitor.domain != result.target and top_competitor.sov > sov_value:
+                recs.append(
+                    f"최대 경쟁사 {top_competitor.domain}의 SOV가 "
+                    f"{top_competitor.sov}%로 앞서고 있습니다. "
+                    "해당 경쟁사의 콘텐츠 전략과 인용 패턴을 분석하세요."
+                )
+
+        # General best practices
+        recs.append(
+            "AI 검색 최적화를 위해 다음 사항을 지속적으로 점검하세요: "
+            "(1) 구조화된 데이터(JSON-LD) 적용, "
+            "(2) FAQ 및 How-to 콘텐츠 발행, "
+            "(3) 신뢰할 수 있는 외부 사이트에서의 백링크 확보, "
+            "(4) 콘텐츠 정기 업데이트 및 정확성 검증."
+        )
+
+        return recs
+
+    # ---- Main Orchestrator ----
+
+    async def track(
+        self,
+        target: str,
+        competitors: list[str] | None = None,
+        include_history: bool = False,
+        include_sov: bool = False,
+    ) -> AiVisibilityResult:
+        """
+        Orchestrate full AI visibility tracking.
+
+        Args:
+            target: Domain to track
+            competitors: Optional list of competitor domains
+            include_history: Whether to fetch historical trends
+            include_sov: Whether to include SOV analysis
+        """
+        self.logger.info(f"Starting AI visibility tracking for {target}")
+        result = AiVisibilityResult(target=target)
+
+        # Core metrics (always fetched)
+        result.impressions = await self.get_impressions_overview(target)
+        result.mentions = await self.get_mentions_overview(target)
+
+        # Share of Voice
+        if include_sov or competitors:
+            result.share_of_voice = await self.get_sov_overview(target)
+
+        # History
+        if include_history:
+            result.impressions_history = await self.get_impressions_history(target)
+            result.mentions_history = await self.get_mentions_history(target)
+            if include_sov:
+                result.sov_history = await self.get_sov_history(target)
+
+        # Competitor comparison
+        if competitors:
+            result.competitors = await self.compare_competitors(target, competitors)
+
+        # Generate recommendations
+        result.recommendations = self.generate_recommendations(result)
+
+        self.print_stats()
+        return result
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def build_parser() -> argparse.ArgumentParser:
+    """Build argument parser for CLI usage."""
+    parser = argparse.ArgumentParser(
+        description="AI Visibility Tracker - Monitor brand visibility in AI search results",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  %(prog)s --target example.com --json
+  %(prog)s --target example.com --competitor comp1.com --competitor comp2.com --json
+  %(prog)s --target example.com --history --sov --json
+  %(prog)s --target example.com --output report.json
+        """,
+    )
+    parser.add_argument(
+        "--target", required=True,
+        help="Target domain to track (e.g., example.com)",
+    )
+    parser.add_argument(
+        "--competitor", action="append", default=[],
+        help="Competitor domain (repeatable). e.g., --competitor a.com --competitor b.com",
+    )
+    parser.add_argument(
+        "--history", action="store_true",
+        help="Include historical trend data (impressions, mentions, SOV over time)",
+    )
+    parser.add_argument(
+        "--sov", action="store_true",
+        help="Include Share of Voice analysis",
+    )
+    parser.add_argument(
+        "--json", action="store_true",
+        help="Output result as JSON to stdout",
+    )
+    parser.add_argument(
+        "--output", type=str, default=None,
+        help="Save JSON output to file path",
+    )
+    return parser
+
+
+def print_summary(result: AiVisibilityResult) -> None:
+    """Print a human-readable summary of AI visibility results."""
+    print("\n" + "=" * 60)
+    print(f"  AI Visibility Report: {result.target}")
+    print("=" * 60)
+
+    print(f"\n  Impressions: {result.impressions.total:,}")
+    print(f"    Trend: {result.impressions.trend} ({result.impressions.change_pct:+.1f}%)")
+
+    print(f"\n  Mentions: {result.mentions.total:,}")
+    print(f"    Trend: {result.mentions.trend} ({result.mentions.change_pct:+.1f}%)")
+
+    if result.share_of_voice:
+        sov = result.share_of_voice.get("brand_sov", 0.0)
+        print(f"\n  Share of Voice: {sov:.1f}%")
+        comp_list = result.share_of_voice.get("competitors", [])
+        if comp_list:
+            print("    Competitors:")
+            for c in comp_list:
+                print(f"      {c.get('domain', '?')}: {c.get('sov_pct', 0):.1f}%")
+
+    if result.impressions_history:
+        trend_info = AiVisibilityTracker.calculate_trends(result.impressions_history)
+        print(f"\n  Impressions Trend: {trend_info['direction']}")
+        print(f"    Range: {trend_info['min_value']:,.0f} - {trend_info['max_value']:,.0f}")
+        print(f"    Change: {trend_info['change_pct']:+.1f}%")
+
+    if result.competitors:
+        print("\n  Competitor Comparison:")
+        for cv in result.competitors:
+            marker = " <-- target" if cv.domain == result.target else ""
+            print(f"    {cv.domain}: SOV={cv.sov:.1f}%, Imp={cv.impressions:,}, Men={cv.mentions:,}{marker}")
+
+    if result.recommendations:
+        print("\n  Recommendations:")
+        for i, rec in enumerate(result.recommendations, 1):
+            print(f"    {i}. {rec}")
+
+    print("\n" + "=" * 60)
+    print(f"  Generated: {result.timestamp}")
+    print("=" * 60 + "\n")
+
+
+async def main() -> None:
+    """CLI entry point."""
+    parser = build_parser()
+    args = parser.parse_args()
+
+    tracker = AiVisibilityTracker(
+        max_concurrent=5,
+        requests_per_second=2.0,
+    )
+
+    result = await tracker.track(
+        target=args.target,
+        competitors=args.competitor if args.competitor else None,
+        include_history=args.history,
+        include_sov=args.sov,
+    )
+
+    # Output
+    if args.json or args.output:
+        output_data = result.to_dict()
+        json_str = json.dumps(output_data, ensure_ascii=False, indent=2)
+
+        if args.json:
+            print(json_str)
+
+        if args.output:
+            output_path = Path(args.output)
+            output_path.parent.mkdir(parents=True, exist_ok=True)
+            output_path.write_text(json_str, encoding="utf-8")
+            logger.info(f"Report saved to {args.output}")
+    else:
+        print_summary(result)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/27-seo-ai-visibility/code/scripts/base_client.py b/custom-skills/27-seo-ai-visibility/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/27-seo-ai-visibility/code/scripts/requirements.txt b/custom-skills/27-seo-ai-visibility/code/scripts/requirements.txt
new file mode 100644
index 0000000..f1881e6
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/code/scripts/requirements.txt
@@ -0,0 +1,8 @@
+# 27-seo-ai-visibility dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+pandas>=2.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/27-seo-ai-visibility/desktop/SKILL.md b/custom-skills/27-seo-ai-visibility/desktop/SKILL.md
new file mode 100644
index 0000000..576baa4
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/desktop/SKILL.md
@@ -0,0 +1,66 @@
+---
+name: seo-ai-visibility
+description: |
+  AI search visibility and brand radar monitoring. Tracks how a brand appears
+  in AI-generated search answers using Ahrefs Brand Radar APIs.
+  Triggers: AI search, AI visibility, brand radar, AI citations,
+  share of voice, AI answers, AI mentions.
+---
+
+# SEO AI Visibility & Brand Radar
+
+Monitor and analyze brand visibility in AI-generated search results. This skill uses Ahrefs Brand Radar APIs to track impressions, mentions, share of voice, cited domains, cited pages, and AI response content.
+
+## Capabilities
+
+### AI Visibility Tracking
+- **Impressions Overview** - How often the brand appears in AI answers
+- **Mentions Overview** - Brand mention frequency across AI engines
+- **Share of Voice (SOV)** - Brand's share vs competitors in AI search
+- **Historical Trends** - Impressions, mentions, and SOV over time
+- **Competitor Comparison** - Side-by-side AI visibility metrics
+
+### AI Citation Analysis
+- **AI Response Analysis** - Content and sentiment of AI mentions
+- **Cited Domains** - Which source domains AI engines reference
+- **Cited Pages** - Specific URLs that get cited in AI answers
+- **Citation Ranking** - Frequency-based ranking of citations
+- **Sentiment Analysis** - Positive/neutral/negative distribution
+
+## Workflow
+
+1. **Input**: User provides target domain and optional competitors
+2. **Data Collection**: Fetch metrics from Ahrefs Brand Radar APIs
+3. **Analysis**: Calculate trends, compare competitors, analyze sentiment
+4. **Recommendations**: Generate actionable Korean-language recommendations
+5. **Output**: JSON report and Notion database entry
+
+## Ahrefs MCP Tools
+
+| Tool | Purpose |
+|------|---------|
+| `brand-radar-ai-responses` | AI-generated responses mentioning brand |
+| `brand-radar-cited-domains` | Domains cited in AI answers |
+| `brand-radar-cited-pages` | Specific pages cited in AI answers |
+| `brand-radar-impressions-history` | Impression trend over time |
+| `brand-radar-impressions-overview` | Current impression metrics |
+| `brand-radar-mentions-history` | Mention trend over time |
+| `brand-radar-mentions-overview` | Current mention metrics |
+| `brand-radar-sov-history` | Share of voice trend |
+| `brand-radar-sov-overview` | Current share of voice |
+
+## Notion Output
+
+All reports are saved to the OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Category**: AI Search Visibility
+- **Audit ID Format**: AI-YYYYMMDD-NNN
+- **Language**: Korean (technical terms in English)
+
+## Example Queries
+
+- "example.com의 AI 검색 가시성을 분석해줘"
+- "AI search visibility for example.com with competitors"
+- "브랜드 레이더 분석: example.com vs competitor.com"
+- "AI 인용 분석 - 어떤 페이지가 AI 답변에서 인용되나요?"
+- "Share of Voice in AI search for our domain"
diff --git a/custom-skills/27-seo-ai-visibility/desktop/skill.yaml b/custom-skills/27-seo-ai-visibility/desktop/skill.yaml
new file mode 100644
index 0000000..a078255
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-ai-visibility
+description: |
+  AI search visibility and brand radar monitoring. Triggers: AI search, AI visibility, brand radar, AI citations, share of voice, AI answers, AI mentions.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/27-seo-ai-visibility/desktop/tools/ahrefs.md b/custom-skills/27-seo-ai-visibility/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..31723b1
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/desktop/tools/ahrefs.md
@@ -0,0 +1,55 @@
+# Ahrefs Brand Radar MCP Tools
+
+## brand-radar-impressions-overview
+Get current AI search impression metrics for a target domain. Returns total impressions, change percentage, and breakdown by AI engine.
+
+**Parameters:**
+- `target` (required): Domain to analyze (e.g., "example.com")
+
+## brand-radar-impressions-history
+Get historical AI search impression data over time. Returns time-series data points with date and impression values.
+
+**Parameters:**
+- `target` (required): Domain to analyze
+
+## brand-radar-mentions-overview
+Get current AI mention metrics for a target domain. Returns total mentions, change percentage, and breakdown.
+
+**Parameters:**
+- `target` (required): Domain to analyze
+
+## brand-radar-mentions-history
+Get historical AI mention data over time. Returns time-series data points with date and mention values.
+
+**Parameters:**
+- `target` (required): Domain to analyze
+
+## brand-radar-sov-overview
+Get Share of Voice overview in AI search for a target domain. Returns brand SOV percentage and competitor SOV data.
+
+**Parameters:**
+- `target` (required): Domain to analyze
+
+## brand-radar-sov-history
+Get historical Share of Voice data over time. Returns time-series SOV data points.
+
+**Parameters:**
+- `target` (required): Domain to analyze
+
+## brand-radar-ai-responses
+Get AI-generated responses that mention the brand. Returns query, response text, sentiment, and source engine for each response.
+
+**Parameters:**
+- `target` (required): Domain to analyze
+
+## brand-radar-cited-domains
+Get domains cited in AI answers related to the brand. Returns domain name, citation count, topics, and share percentage.
+
+**Parameters:**
+- `target` (required): Domain to analyze
+
+## brand-radar-cited-pages
+Get specific pages cited in AI answers. Returns URL, title, citation count, context snippet, and topics.
+
+**Parameters:**
+- `target` (required): Domain to analyze
diff --git a/custom-skills/27-seo-ai-visibility/desktop/tools/notion.md b/custom-skills/27-seo-ai-visibility/desktop/tools/notion.md
new file mode 100644
index 0000000..88a8cb0
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/desktop/tools/notion.md
@@ -0,0 +1,44 @@
+# Notion MCP Tools
+
+## Database: OurDigital SEO Audit Log
+
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **URL**: https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef
+
+## Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title in Korean + date |
+| Site | URL | Tracked website URL |
+| Category | Select | "AI Search Visibility" |
+| Priority | Select | Based on SOV trend (Critical, High, Medium, Low) |
+| Found Date | Date | Report date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: AI-YYYYMMDD-NNN |
+
+## Usage
+
+Use `notion-create-pages` to save audit results:
+
+```json
+{
+  "parent": {"database_id": "2c8581e5-8a1e-8035-880b-e38cefc2f3ef"},
+  "properties": {
+    "Issue": {"title": [{"text": {"content": "AI 검색 가시성 분석 - example.com (2025-01-15)"}}]},
+    "Site": {"url": "https://example.com"},
+    "Category": {"select": {"name": "AI Search Visibility"}},
+    "Priority": {"select": {"name": "Medium"}},
+    "Found Date": {"date": {"start": "2025-01-15"}},
+    "Audit ID": {"rich_text": [{"text": {"content": "AI-20250115-001"}}]}
+  }
+}
+```
+
+## Priority Guidelines
+
+| Condition | Priority |
+|-----------|----------|
+| SOV decreasing >10% | Critical |
+| SOV decreasing 3-10% | High |
+| SOV stable, low (<10%) | Medium |
+| SOV increasing or high (>25%) | Low |
diff --git a/custom-skills/27-seo-ai-visibility/desktop/tools/websearch.md b/custom-skills/27-seo-ai-visibility/desktop/tools/websearch.md
new file mode 100644
index 0000000..9697cdb
--- /dev/null
+++ b/custom-skills/27-seo-ai-visibility/desktop/tools/websearch.md
@@ -0,0 +1,17 @@
+# WebSearch & WebFetch Tools
+
+## WebSearch
+
+Use web search to supplement AI visibility analysis with additional context:
+- Research competitor AI optimization strategies
+- Find industry benchmarks for AI search visibility
+- Look up latest AI search engine algorithm updates
+- Discover best practices for AI citation optimization
+
+## WebFetch
+
+Use web fetch to retrieve specific pages for deeper analysis:
+- Fetch competitor pages that are frequently cited in AI answers
+- Retrieve structured data (Schema Markup) from cited pages
+- Analyze content structure of top-cited URLs
+- Check E-E-A-T signals on referenced pages
diff --git a/custom-skills/28-seo-knowledge-graph/code/CLAUDE.md b/custom-skills/28-seo-knowledge-graph/code/CLAUDE.md
new file mode 100644
index 0000000..9c686f9
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/code/CLAUDE.md
@@ -0,0 +1,139 @@
+# CLAUDE.md
+
+## Overview
+
+Knowledge Graph and Entity SEO tool for analyzing brand entity presence in Google Knowledge Graph, Knowledge Panels, People Also Ask (PAA), and FAQ rich results. Checks entity attribute completeness, Wikipedia/Wikidata presence, and Korean equivalents (Naver knowledge iN, Naver encyclopedia). Uses WebSearch and WebFetch for data collection, Ahrefs serp-overview for SERP feature detection.
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Knowledge Graph analysis
+python scripts/knowledge_graph_analyzer.py --entity "Samsung Electronics" --json
+
+# Entity SEO audit
+python scripts/entity_auditor.py --url https://example.com --entity "Brand Name" --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `knowledge_graph_analyzer.py` | Analyze Knowledge Panel and entity presence | KP detection, entity attributes, Wikipedia/Wikidata status |
+| `entity_auditor.py` | Audit entity SEO signals and PAA/FAQ presence | PAA monitoring, FAQ schema tracking, entity completeness |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## Knowledge Graph Analyzer
+
+```bash
+# Analyze entity in Knowledge Graph
+python scripts/knowledge_graph_analyzer.py --entity "Samsung Electronics" --json
+
+# Check with Korean name
+python scripts/knowledge_graph_analyzer.py --entity "삼성전자" --language ko --json
+
+# Include Wikipedia/Wikidata check
+python scripts/knowledge_graph_analyzer.py --entity "Samsung" --wiki --json
+```
+
+**Capabilities**:
+- Knowledge Panel detection via Google search
+- Entity attribute extraction (name, description, logo, type, social profiles, website)
+- Entity attribute completeness scoring
+- Wikipedia article presence check
+- Wikidata entity presence check (QID lookup)
+- Naver encyclopedia (네이버 백과사전) presence
+- Naver knowledge iN (지식iN) presence
+- Knowledge Panel comparison with competitors
+
+## Entity Auditor
+
+```bash
+# Full entity SEO audit
+python scripts/entity_auditor.py --url https://example.com --entity "Brand Name" --json
+
+# PAA monitoring for brand keywords
+python scripts/entity_auditor.py --url https://example.com --entity "Brand Name" --paa --json
+
+# FAQ rich result tracking
+python scripts/entity_auditor.py --url https://example.com --entity "Brand Name" --faq --json
+```
+
+**Capabilities**:
+- People Also Ask (PAA) monitoring for brand-related queries
+- FAQ schema presence tracking (FAQPage schema -> SERP appearance)
+- Entity markup audit (Organization, Person, LocalBusiness schema on website)
+- Social profile linking validation (sameAs in schema)
+- Brand SERP analysis (what appears when you search the brand name)
+- Entity consistency across web properties
+- Korean entity optimization (Korean Knowledge Panel, Naver profiles)
+
+## Data Sources
+
+| Source | Purpose |
+|--------|---------|
+| WebSearch | Search for entity/brand to detect Knowledge Panel |
+| WebFetch | Fetch Wikipedia, Wikidata, Naver pages |
+| Ahrefs `serp-overview` | SERP feature detection for entity keywords |
+
+## Output Format
+
+```json
+{
+  "entity": "Samsung Electronics",
+  "knowledge_panel": {
+    "detected": true,
+    "attributes": {
+      "name": "Samsung Electronics",
+      "type": "Corporation",
+      "description": "...",
+      "logo": true,
+      "website": true,
+      "social_profiles": ["twitter", "facebook", "linkedin"]
+    },
+    "completeness_score": 85
+  },
+  "wikipedia": {"present": true, "url": "..."},
+  "wikidata": {"present": true, "qid": "Q20710"},
+  "naver_encyclopedia": {"present": true, "url": "..."},
+  "naver_knowledge_in": {"present": true, "entries": 15},
+  "paa_questions": [...],
+  "faq_rich_results": [...],
+  "entity_schema_on_site": {
+    "organization": true,
+    "same_as_links": 5,
+    "completeness": 78
+  },
+  "score": 75,
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Entity website URL |
+| Category | Select | Knowledge Graph & Entity SEO |
+| Priority | Select | Based on entity completeness |
+| Found Date | Date | Audit date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: KG-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., Knowledge Panel, Knowledge Graph, PAA)
+- URLs and code remain unchanged
diff --git a/custom-skills/28-seo-knowledge-graph/code/scripts/base_client.py b/custom-skills/28-seo-knowledge-graph/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/28-seo-knowledge-graph/code/scripts/entity_auditor.py b/custom-skills/28-seo-knowledge-graph/code/scripts/entity_auditor.py
new file mode 100644
index 0000000..63231b5
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/code/scripts/entity_auditor.py
@@ -0,0 +1,902 @@
+"""
+Entity Auditor
+===============
+Purpose: Audit entity SEO signals including PAA monitoring, FAQ schema tracking,
+         entity markup validation, and brand SERP analysis.
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import re
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from typing import Any
+from urllib.parse import quote, urljoin, urlparse
+
+import aiohttp
+from bs4 import BeautifulSoup
+from rich.console import Console
+from rich.table import Table
+
+from base_client import BaseAsyncClient, ConfigManager, config
+
+logger = logging.getLogger(__name__)
+console = Console()
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+
+@dataclass
+class PaaQuestion:
+    """A People Also Ask question found in SERP."""
+    question: str = ""
+    keyword: str = ""
+    position: int = 0
+    source_url: str | None = None
+
+
+@dataclass
+class FaqRichResult:
+    """FAQ rich result tracking entry."""
+    url: str = ""
+    question_count: int = 0
+    appearing_in_serp: bool = False
+    questions: list[str] = field(default_factory=list)
+    schema_valid: bool = False
+
+
+@dataclass
+class EntitySchema:
+    """Entity structured data found on a website."""
+    type: str = ""  # Organization, Person, LocalBusiness, etc.
+    properties: dict[str, Any] = field(default_factory=dict)
+    same_as_links: list[str] = field(default_factory=list)
+    completeness: float = 0.0
+    issues: list[str] = field(default_factory=list)
+
+
+@dataclass
+class BrandSerpResult:
+    """What appears when searching for the brand name."""
+    query: str = ""
+    features: list[str] = field(default_factory=list)
+    paa_count: int = 0
+    faq_count: int = 0
+    knowledge_panel: bool = False
+    sitelinks: bool = False
+    social_profiles: list[str] = field(default_factory=list)
+    top_results: list[dict[str, str]] = field(default_factory=list)
+
+
+@dataclass
+class EntityAuditResult:
+    """Full entity SEO audit result."""
+    url: str = ""
+    entity_name: str = ""
+    paa_questions: list[PaaQuestion] = field(default_factory=list)
+    faq_rich_results: list[FaqRichResult] = field(default_factory=list)
+    entity_schemas: list[EntitySchema] = field(default_factory=list)
+    brand_serp: BrandSerpResult = field(default_factory=BrandSerpResult)
+    social_profile_status: dict[str, bool] = field(default_factory=dict)
+    overall_score: float = 0.0
+    recommendations: list[str] = field(default_factory=list)
+    timestamp: str = field(default_factory=lambda: datetime.now().isoformat())
+
+    def to_dict(self) -> dict[str, Any]:
+        return asdict(self)
+
+
+# ---------------------------------------------------------------------------
+# Entity Auditor
+# ---------------------------------------------------------------------------
+
+
+class EntityAuditor(BaseAsyncClient):
+    """Audit entity SEO signals and rich result presence."""
+
+    GOOGLE_SEARCH_URL = "https://www.google.com/search"
+
+    HEADERS = {
+        "User-Agent": (
+            "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
+            "AppleWebKit/537.36 (KHTML, like Gecko) "
+            "Chrome/120.0.0.0 Safari/537.36"
+        ),
+        "Accept-Language": "en-US,en;q=0.9",
+    }
+
+    PAA_KEYWORD_TEMPLATES = [
+        "{entity}",
+        "{entity} reviews",
+        "{entity} vs",
+        "what is {entity}",
+        "{entity} pricing",
+        "{entity} alternatives",
+        "is {entity} good",
+        "{entity} benefits",
+        "how to use {entity}",
+        "{entity} complaints",
+    ]
+
+    EXPECTED_SCHEMA_PROPERTIES = {
+        "Organization": [
+            "name", "url", "logo", "description", "sameAs",
+            "contactPoint", "address", "foundingDate", "founder",
+            "numberOfEmployees", "email", "telephone",
+        ],
+        "Person": [
+            "name", "url", "image", "description", "sameAs",
+            "jobTitle", "worksFor", "alumniOf", "birthDate",
+        ],
+        "LocalBusiness": [
+            "name", "url", "image", "description", "sameAs",
+            "address", "telephone", "openingHours", "geo",
+            "priceRange", "aggregateRating",
+        ],
+    }
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.config = config
+
+    # ------------------------------------------------------------------
+    # PAA monitoring
+    # ------------------------------------------------------------------
+
+    async def monitor_paa(
+        self,
+        entity_name: str,
+        keywords: list[str] | None = None,
+        session: aiohttp.ClientSession | None = None,
+    ) -> list[PaaQuestion]:
+        """Search brand keywords and extract People Also Ask questions."""
+        if keywords is None:
+            keywords = [t.format(entity=entity_name) for t in self.PAA_KEYWORD_TEMPLATES]
+
+        paa_questions: list[PaaQuestion] = []
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            for keyword in keywords:
+                params = {"q": keyword, "hl": "en", "gl": "us"}
+                try:
+                    async with session.get(
+                        self.GOOGLE_SEARCH_URL, params=params, headers=self.HEADERS,
+                        timeout=aiohttp.ClientTimeout(total=20),
+                    ) as resp:
+                        if resp.status != 200:
+                            logger.warning("Search for '%s' returned status %d", keyword, resp.status)
+                            continue
+
+                        html = await resp.text()
+                        soup = BeautifulSoup(html, "lxml")
+
+                        # PAA box selectors
+                        paa_selectors = [
+                            "div[data-sgrd] div[data-q]",
+                            "div.related-question-pair",
+                            "div[jsname] div[data-q]",
+                            "div.wQiwMc",
+                        ]
+
+                        position = 0
+                        for selector in paa_selectors:
+                            elements = soup.select(selector)
+                            for el in elements:
+                                question_text = el.get("data-q", "") or el.get_text(strip=True)
+                                if question_text and len(question_text) > 5:
+                                    position += 1
+                                    paa_questions.append(PaaQuestion(
+                                        question=question_text,
+                                        keyword=keyword,
+                                        position=position,
+                                    ))
+
+                        # Fallback: regex for PAA-like questions
+                        if not paa_questions:
+                            text = soup.get_text(separator="\n")
+                            q_patterns = re.findall(
+                                r"((?:What|How|Why|When|Where|Who|Is|Can|Does|Do|Which)\s+[^?\n]{10,80}\??)",
+                                text,
+                            )
+                            for i, q in enumerate(q_patterns[:8]):
+                                paa_questions.append(PaaQuestion(
+                                    question=q.strip(),
+                                    keyword=keyword,
+                                    position=i + 1,
+                                ))
+
+                except Exception as exc:
+                    logger.error("PAA search failed for '%s': %s", keyword, exc)
+                    continue
+
+                # Rate limit between searches
+                await asyncio.sleep(1.5)
+        finally:
+            if own_session:
+                await session.close()
+
+        # Deduplicate questions
+        seen = set()
+        unique = []
+        for q in paa_questions:
+            key = q.question.lower().strip()
+            if key not in seen:
+                seen.add(key)
+                unique.append(q)
+
+        logger.info("Found %d unique PAA questions for '%s'", len(unique), entity_name)
+        return unique
+
+    # ------------------------------------------------------------------
+    # FAQ rich result tracking
+    # ------------------------------------------------------------------
+
+    async def track_faq_rich_results(
+        self,
+        url: str,
+        session: aiohttp.ClientSession | None = None,
+    ) -> list[FaqRichResult]:
+        """Check pages for FAQPage schema and SERP appearance."""
+        faq_results: list[FaqRichResult] = []
+        domain = urlparse(url).netloc
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            # Fetch the page and look for FAQ schema
+            async with session.get(
+                url, headers=self.HEADERS, timeout=aiohttp.ClientTimeout(total=20),
+            ) as resp:
+                if resp.status != 200:
+                    logger.warning("Page %s returned status %d", url, resp.status)
+                    return faq_results
+
+                html = await resp.text()
+                soup = BeautifulSoup(html, "lxml")
+
+                # Find JSON-LD scripts with FAQPage
+                scripts = soup.find_all("script", type="application/ld+json")
+                for script in scripts:
+                    try:
+                        data = json.loads(script.string or "{}")
+                        items = data if isinstance(data, list) else [data]
+
+                        for item in items:
+                            schema_type = item.get("@type", "")
+                            if schema_type == "FAQPage" or (
+                                isinstance(schema_type, list) and "FAQPage" in schema_type
+                            ):
+                                questions = item.get("mainEntity", [])
+                                faq = FaqRichResult(
+                                    url=url,
+                                    question_count=len(questions),
+                                    questions=[
+                                        q.get("name", "") for q in questions if isinstance(q, dict)
+                                    ],
+                                    schema_valid=True,
+                                )
+                                faq_results.append(faq)
+
+                            # Check for nested @graph
+                            graph = item.get("@graph", [])
+                            for g_item in graph:
+                                if g_item.get("@type") == "FAQPage":
+                                    questions = g_item.get("mainEntity", [])
+                                    faq = FaqRichResult(
+                                        url=url,
+                                        question_count=len(questions),
+                                        questions=[
+                                            q.get("name", "") for q in questions if isinstance(q, dict)
+                                        ],
+                                        schema_valid=True,
+                                    )
+                                    faq_results.append(faq)
+
+                    except json.JSONDecodeError:
+                        continue
+
+                # Also check for microdata FAQ markup
+                faq_items = soup.select("[itemtype*='FAQPage'] [itemprop='mainEntity']")
+                if faq_items and not faq_results:
+                    questions = []
+                    for item in faq_items:
+                        q_el = item.select_one("[itemprop='name']")
+                        if q_el:
+                            questions.append(q_el.get_text(strip=True))
+                    faq_results.append(FaqRichResult(
+                        url=url,
+                        question_count=len(questions),
+                        questions=questions,
+                        schema_valid=True,
+                    ))
+
+        except Exception as exc:
+            logger.error("FAQ tracking failed for %s: %s", url, exc)
+        finally:
+            if own_session:
+                await session.close()
+
+        logger.info("Found %d FAQ schemas on %s", len(faq_results), url)
+        return faq_results
+
+    # ------------------------------------------------------------------
+    # Entity schema audit
+    # ------------------------------------------------------------------
+
+    async def audit_entity_schema(
+        self,
+        url: str,
+        session: aiohttp.ClientSession | None = None,
+    ) -> list[EntitySchema]:
+        """Check Organization/Person/LocalBusiness schema on website."""
+        schemas: list[EntitySchema] = []
+        target_types = {"Organization", "Person", "LocalBusiness", "Corporation", "MedicalBusiness"}
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            async with session.get(
+                url, headers=self.HEADERS, timeout=aiohttp.ClientTimeout(total=20),
+            ) as resp:
+                if resp.status != 200:
+                    logger.warning("Page %s returned status %d", url, resp.status)
+                    return schemas
+
+                html = await resp.text()
+                soup = BeautifulSoup(html, "lxml")
+
+                scripts = soup.find_all("script", type="application/ld+json")
+                for script in scripts:
+                    try:
+                        data = json.loads(script.string or "{}")
+                        items = data if isinstance(data, list) else [data]
+
+                        # Include @graph nested items
+                        expanded = []
+                        for item in items:
+                            expanded.append(item)
+                            if "@graph" in item:
+                                expanded.extend(item["@graph"])
+
+                        for item in expanded:
+                            item_type = item.get("@type", "")
+                            if isinstance(item_type, list):
+                                matching = [t for t in item_type if t in target_types]
+                                if not matching:
+                                    continue
+                                item_type = matching[0]
+                            elif item_type not in target_types:
+                                continue
+
+                            same_as = item.get("sameAs", [])
+                            if isinstance(same_as, str):
+                                same_as = [same_as]
+
+                            # Calculate completeness
+                            base_type = item_type
+                            if base_type == "Corporation":
+                                base_type = "Organization"
+                            elif base_type == "MedicalBusiness":
+                                base_type = "LocalBusiness"
+
+                            expected = self.EXPECTED_SCHEMA_PROPERTIES.get(base_type, [])
+                            present = [k for k in expected if k in item and item[k]]
+                            completeness = round((len(present) / len(expected)) * 100, 1) if expected else 0
+
+                            # Check for issues
+                            issues = []
+                            if "name" not in item:
+                                issues.append("Missing 'name' property")
+                            if "url" not in item:
+                                issues.append("Missing 'url' property")
+                            if not same_as:
+                                issues.append("No 'sameAs' links (social profiles)")
+                            if "logo" not in item and base_type == "Organization":
+                                issues.append("Missing 'logo' property")
+                            if "description" not in item:
+                                issues.append("Missing 'description' property")
+
+                            schema = EntitySchema(
+                                type=item_type,
+                                properties={k: (str(v)[:100] if not isinstance(v, (list, dict)) else v) for k, v in item.items() if k != "@context"},
+                                same_as_links=same_as,
+                                completeness=completeness,
+                                issues=issues,
+                            )
+                            schemas.append(schema)
+
+                    except json.JSONDecodeError:
+                        continue
+
+        except Exception as exc:
+            logger.error("Entity schema audit failed for %s: %s", url, exc)
+        finally:
+            if own_session:
+                await session.close()
+
+        logger.info("Found %d entity schemas on %s", len(schemas), url)
+        return schemas
+
+    # ------------------------------------------------------------------
+    # Brand SERP analysis
+    # ------------------------------------------------------------------
+
+    async def analyze_brand_serp(
+        self,
+        entity_name: str,
+        session: aiohttp.ClientSession | None = None,
+    ) -> BrandSerpResult:
+        """Analyze what appears in SERP for the brand name search."""
+        result = BrandSerpResult(query=entity_name)
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            params = {"q": entity_name, "hl": "en", "gl": "us"}
+            async with session.get(
+                self.GOOGLE_SEARCH_URL, params=params, headers=self.HEADERS,
+                timeout=aiohttp.ClientTimeout(total=20),
+            ) as resp:
+                if resp.status != 200:
+                    return result
+
+                html = await resp.text()
+                soup = BeautifulSoup(html, "lxml")
+                text = soup.get_text(separator=" ", strip=True).lower()
+
+                # Detect SERP features
+                feature_indicators = {
+                    "knowledge_panel": ["kp-wholepage", "knowledge-panel", "kno-"],
+                    "sitelinks": ["sitelinks", "site-links"],
+                    "people_also_ask": ["related-question-pair", "data-q"],
+                    "faq_rich_result": ["faqpage", "frequently asked"],
+                    "featured_snippet": ["featured-snippet", "data-tts"],
+                    "image_pack": ["image-result", "img-brk"],
+                    "video_carousel": ["video-result", "vid-"],
+                    "twitter_carousel": ["twitter-timeline", "g-scrolling-carousel"],
+                    "reviews": ["star-rating", "aggregate-rating"],
+                    "local_pack": ["local-pack", "local_pack"],
+                }
+
+                for feature, indicators in feature_indicators.items():
+                    for ind in indicators:
+                        if ind in str(soup).lower():
+                            result.features.append(feature)
+                            break
+
+                result.knowledge_panel = "knowledge_panel" in result.features
+                result.sitelinks = "sitelinks" in result.features
+
+                # Count PAA questions
+                paa_elements = soup.select("div[data-q], div.related-question-pair")
+                result.paa_count = len(paa_elements)
+                if result.paa_count > 0 and "people_also_ask" not in result.features:
+                    result.features.append("people_also_ask")
+
+                # Detect social profiles in results
+                social_domains = {
+                    "twitter.com": "twitter", "x.com": "twitter",
+                    "facebook.com": "facebook", "linkedin.com": "linkedin",
+                    "youtube.com": "youtube", "instagram.com": "instagram",
+                    "github.com": "github", "pinterest.com": "pinterest",
+                }
+                links = soup.find_all("a", href=True)
+                for link in links:
+                    href = link["href"]
+                    for domain, name in social_domains.items():
+                        if domain in href and name not in result.social_profiles:
+                            result.social_profiles.append(name)
+
+                # Extract top organic results
+                result_divs = soup.select("div.g, div[data-sokoban-container]")[:10]
+                for div in result_divs:
+                    title_el = div.select_one("h3")
+                    link_el = div.select_one("a[href]")
+                    if title_el and link_el:
+                        result.top_results.append({
+                            "title": title_el.get_text(strip=True),
+                            "url": link_el.get("href", ""),
+                        })
+
+        except Exception as exc:
+            logger.error("Brand SERP analysis failed for '%s': %s", entity_name, exc)
+        finally:
+            if own_session:
+                await session.close()
+
+        return result
+
+    # ------------------------------------------------------------------
+    # Social profile link validation
+    # ------------------------------------------------------------------
+
+    async def check_social_profile_links(
+        self,
+        same_as_links: list[str],
+        session: aiohttp.ClientSession | None = None,
+    ) -> dict[str, bool]:
+        """Validate sameAs URLs are accessible."""
+        status: dict[str, bool] = {}
+        if not same_as_links:
+            return status
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            for link in same_as_links:
+                try:
+                    async with session.head(
+                        link, headers=self.HEADERS, timeout=aiohttp.ClientTimeout(total=10),
+                        allow_redirects=True,
+                    ) as resp:
+                        status[link] = resp.status < 400
+                except Exception:
+                    status[link] = False
+
+                await asyncio.sleep(0.5)
+        finally:
+            if own_session:
+                await session.close()
+
+        accessible = sum(1 for v in status.values() if v)
+        logger.info("Social profile links: %d/%d accessible", accessible, len(status))
+        return status
+
+    # ------------------------------------------------------------------
+    # Recommendations
+    # ------------------------------------------------------------------
+
+    def generate_recommendations(self, result: EntityAuditResult) -> list[str]:
+        """Generate actionable entity SEO improvement recommendations."""
+        recs: list[str] = []
+
+        # PAA recommendations
+        if not result.paa_questions:
+            recs.append(
+                "브랜드 관련 People Also Ask(PAA) 질문이 감지되지 않았습니다. "
+                "FAQ 콘텐츠를 작성하여 PAA 노출 기회를 확보하세요."
+            )
+        elif len(result.paa_questions) < 5:
+            recs.append(
+                f"PAA 질문이 {len(result.paa_questions)}개만 감지되었습니다. "
+                "더 다양한 키워드에 대한 Q&A 콘텐츠를 강화하세요."
+            )
+
+        # FAQ schema recommendations
+        if not result.faq_rich_results:
+            recs.append(
+                "FAQPage schema가 감지되지 않았습니다. "
+                "FAQ 페이지에 FAQPage JSON-LD를 추가하여 Rich Result를 확보하세요."
+            )
+        else:
+            invalid = [f for f in result.faq_rich_results if not f.schema_valid]
+            if invalid:
+                recs.append(
+                    f"{len(invalid)}개의 FAQ schema에 유효성 문제가 있습니다. "
+                    "Google Rich Results Test로 검증하세요."
+                )
+
+        # Entity schema recommendations
+        if not result.entity_schemas:
+            recs.append(
+                "Organization/Person/LocalBusiness schema가 없습니다. "
+                "홈페이지에 Organization schema JSON-LD를 추가하세요."
+            )
+        else:
+            for schema in result.entity_schemas:
+                if schema.completeness < 50:
+                    recs.append(
+                        f"{schema.type} schema 완성도가 {schema.completeness}%입니다. "
+                        f"누락 항목: {', '.join(schema.issues[:3])}"
+                    )
+                if not schema.same_as_links:
+                    recs.append(
+                        f"{schema.type} schema에 sameAs 속성이 없습니다. "
+                        "소셜 미디어 프로필 URL을 sameAs에 추가하세요."
+                    )
+
+        # Brand SERP recommendations
+        serp = result.brand_serp
+        if not serp.knowledge_panel:
+            recs.append(
+                "브랜드 검색 시 Knowledge Panel이 표시되지 않습니다. "
+                "Wikipedia, Wikidata, 구조화된 데이터를 통해 엔티티 인식을 강화하세요."
+            )
+        if not serp.sitelinks:
+            recs.append(
+                "Sitelinks가 표시되지 않습니다. "
+                "사이트 구조와 내부 링크를 개선하세요."
+            )
+        if len(serp.social_profiles) < 3:
+            recs.append(
+                f"SERP에 소셜 프로필이 {len(serp.social_profiles)}개만 표시됩니다. "
+                "주요 소셜 미디어 프로필을 활성화하고 schema sameAs에 연결하세요."
+            )
+
+        # Social profile accessibility
+        broken = [url for url, ok in result.social_profile_status.items() if not ok]
+        if broken:
+            recs.append(
+                f"접근 불가한 소셜 프로필 링크 {len(broken)}개: "
+                f"{', '.join(broken[:3])}. sameAs URL을 업데이트하세요."
+            )
+
+        if not recs:
+            recs.append("Entity SEO 상태가 양호합니다. 현재 수준을 유지하세요.")
+
+        return recs
+
+    # ------------------------------------------------------------------
+    # Scoring
+    # ------------------------------------------------------------------
+
+    def compute_score(self, result: EntityAuditResult) -> float:
+        """Compute overall entity SEO score (0-100)."""
+        score = 0.0
+
+        # PAA presence (15 points)
+        paa_count = len(result.paa_questions)
+        if paa_count >= 10:
+            score += 15
+        elif paa_count >= 5:
+            score += 10
+        elif paa_count > 0:
+            score += 5
+
+        # FAQ schema (15 points)
+        if result.faq_rich_results:
+            valid_count = sum(1 for f in result.faq_rich_results if f.schema_valid)
+            score += min(15, valid_count * 5)
+
+        # Entity schema (25 points)
+        if result.entity_schemas:
+            best_completeness = max(s.completeness for s in result.entity_schemas)
+            score += best_completeness * 0.25
+
+        # Brand SERP features (25 points)
+        serp = result.brand_serp
+        if serp.knowledge_panel:
+            score += 10
+        if serp.sitelinks:
+            score += 5
+        score += min(10, len(serp.features) * 2)
+
+        # Social profiles (10 points)
+        if result.social_profile_status:
+            accessible = sum(1 for v in result.social_profile_status.values() if v)
+            total = len(result.social_profile_status)
+            score += (accessible / total) * 10 if total > 0 else 0
+
+        # sameAs links (10 points)
+        total_same_as = sum(len(s.same_as_links) for s in result.entity_schemas)
+        score += min(10, total_same_as * 2)
+
+        return round(min(100, score), 1)
+
+    # ------------------------------------------------------------------
+    # Main orchestrator
+    # ------------------------------------------------------------------
+
+    async def audit(
+        self,
+        url: str,
+        entity_name: str,
+        include_paa: bool = True,
+        include_faq: bool = True,
+    ) -> EntityAuditResult:
+        """Orchestrate full entity SEO audit."""
+        result = EntityAuditResult(url=url, entity_name=entity_name)
+        logger.info("Starting entity audit for '%s' at %s", entity_name, url)
+
+        async with aiohttp.ClientSession() as session:
+            # Parallel tasks: entity schema, brand SERP, FAQ
+            tasks = [
+                self.audit_entity_schema(url, session),
+                self.analyze_brand_serp(entity_name, session),
+            ]
+
+            if include_faq:
+                tasks.append(self.track_faq_rich_results(url, session))
+
+            results = await asyncio.gather(*tasks, return_exceptions=True)
+
+            # Unpack results
+            if not isinstance(results[0], Exception):
+                result.entity_schemas = results[0]
+            else:
+                logger.error("Entity schema audit failed: %s", results[0])
+
+            if not isinstance(results[1], Exception):
+                result.brand_serp = results[1]
+            else:
+                logger.error("Brand SERP analysis failed: %s", results[1])
+
+            if include_faq and len(results) > 2 and not isinstance(results[2], Exception):
+                result.faq_rich_results = results[2]
+
+            # PAA monitoring (sequential due to rate limits)
+            if include_paa:
+                result.paa_questions = await self.monitor_paa(entity_name, session=session)
+
+            # Validate social profile links from schema
+            all_same_as = []
+            for schema in result.entity_schemas:
+                all_same_as.extend(schema.same_as_links)
+            if all_same_as:
+                result.social_profile_status = await self.check_social_profile_links(
+                    list(set(all_same_as)), session
+                )
+
+        # Compute score and recommendations
+        result.overall_score = self.compute_score(result)
+        result.recommendations = self.generate_recommendations(result)
+
+        logger.info("Entity audit complete. Score: %.1f", result.overall_score)
+        return result
+
+
+# ---------------------------------------------------------------------------
+# CLI display helpers
+# ---------------------------------------------------------------------------
+
+
+def display_result(result: EntityAuditResult) -> None:
+    """Display audit result in rich tables."""
+    console.print()
+    console.print(f"[bold cyan]Entity SEO Audit: {result.entity_name}[/bold cyan]")
+    console.print(f"URL: {result.url} | Score: {result.overall_score}/100")
+    console.print()
+
+    # Entity Schema table
+    if result.entity_schemas:
+        table = Table(title="Entity Schema Markup", show_header=True)
+        table.add_column("Type", style="bold")
+        table.add_column("Completeness")
+        table.add_column("sameAs Links")
+        table.add_column("Issues")
+
+        for schema in result.entity_schemas:
+            issues_text = "; ".join(schema.issues[:3]) if schema.issues else "None"
+            table.add_row(
+                schema.type,
+                f"{schema.completeness}%",
+                str(len(schema.same_as_links)),
+                issues_text,
+            )
+        console.print(table)
+    else:
+        console.print("[red]No entity schema markup found on website![/red]")
+    console.print()
+
+    # Brand SERP table
+    serp = result.brand_serp
+    serp_table = Table(title="Brand SERP Analysis", show_header=True)
+    serp_table.add_column("Feature", style="bold")
+    serp_table.add_column("Status")
+
+    serp_table.add_row("Knowledge Panel", "[green]Yes[/]" if serp.knowledge_panel else "[red]No[/]")
+    serp_table.add_row("Sitelinks", "[green]Yes[/]" if serp.sitelinks else "[red]No[/]")
+    serp_table.add_row("PAA Count", str(serp.paa_count))
+    serp_table.add_row("SERP Features", ", ".join(serp.features) if serp.features else "None")
+    serp_table.add_row("Social Profiles", ", ".join(serp.social_profiles) if serp.social_profiles else "None")
+
+    console.print(serp_table)
+    console.print()
+
+    # PAA Questions
+    if result.paa_questions:
+        paa_table = Table(title=f"People Also Ask ({len(result.paa_questions)} questions)", show_header=True)
+        paa_table.add_column("#", style="dim")
+        paa_table.add_column("Question")
+        paa_table.add_column("Keyword")
+
+        for i, q in enumerate(result.paa_questions[:15], 1):
+            paa_table.add_row(str(i), q.question, q.keyword)
+        console.print(paa_table)
+    console.print()
+
+    # FAQ Rich Results
+    if result.faq_rich_results:
+        faq_table = Table(title="FAQ Rich Results", show_header=True)
+        faq_table.add_column("URL")
+        faq_table.add_column("Questions")
+        faq_table.add_column("Valid")
+
+        for faq in result.faq_rich_results:
+            faq_table.add_row(
+                faq.url[:60],
+                str(faq.question_count),
+                "[green]Yes[/]" if faq.schema_valid else "[red]No[/]",
+            )
+        console.print(faq_table)
+    console.print()
+
+    # Social Profile Status
+    if result.social_profile_status:
+        sp_table = Table(title="Social Profile Link Status", show_header=True)
+        sp_table.add_column("URL")
+        sp_table.add_column("Accessible")
+
+        for link, accessible in result.social_profile_status.items():
+            sp_table.add_row(
+                link[:70],
+                "[green]Yes[/]" if accessible else "[red]No[/]",
+            )
+        console.print(sp_table)
+    console.print()
+
+    # Recommendations
+    console.print("[bold yellow]Recommendations:[/bold yellow]")
+    for i, rec in enumerate(result.recommendations, 1):
+        console.print(f"  {i}. {rec}")
+    console.print()
+
+
+# ---------------------------------------------------------------------------
+# Main
+# ---------------------------------------------------------------------------
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Entity SEO Auditor",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    parser.add_argument("--url", required=True, help="Website URL to audit")
+    parser.add_argument("--entity", required=True, help="Entity/brand name")
+    parser.add_argument("--paa", action="store_true", default=True, help="Include PAA monitoring (default: True)")
+    parser.add_argument("--no-paa", action="store_true", help="Skip PAA monitoring")
+    parser.add_argument("--faq", action="store_true", default=True, help="Include FAQ tracking (default: True)")
+    parser.add_argument("--no-faq", action="store_true", help="Skip FAQ tracking")
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", type=str, help="Output file path")
+    return parser.parse_args()
+
+
+async def main() -> None:
+    args = parse_args()
+
+    auditor = EntityAuditor()
+    result = await auditor.audit(
+        url=args.url,
+        entity_name=args.entity,
+        include_paa=not args.no_paa,
+        include_faq=not args.no_faq,
+    )
+
+    if args.json:
+        output = json.dumps(result.to_dict(), ensure_ascii=False, indent=2)
+        if args.output:
+            with open(args.output, "w", encoding="utf-8") as f:
+                f.write(output)
+            console.print(f"[green]Output saved to {args.output}[/green]")
+        else:
+            print(output)
+    else:
+        display_result(result)
+        if args.output:
+            with open(args.output, "w", encoding="utf-8") as f:
+                json.dump(result.to_dict(), f, ensure_ascii=False, indent=2)
+            console.print(f"[green]Output saved to {args.output}[/green]")
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/28-seo-knowledge-graph/code/scripts/knowledge_graph_analyzer.py b/custom-skills/28-seo-knowledge-graph/code/scripts/knowledge_graph_analyzer.py
new file mode 100644
index 0000000..4d371a6
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/code/scripts/knowledge_graph_analyzer.py
@@ -0,0 +1,782 @@
+"""
+Knowledge Graph Analyzer
+=========================
+Purpose: Analyze entity presence in Google Knowledge Graph, Knowledge Panels,
+         Wikipedia, Wikidata, and Korean equivalents (Naver encyclopedia, 지식iN).
+Python: 3.10+
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import re
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from typing import Any
+from urllib.parse import quote, urljoin
+
+import aiohttp
+from bs4 import BeautifulSoup
+from rich.console import Console
+from rich.table import Table
+
+from base_client import BaseAsyncClient, ConfigManager, config
+
+logger = logging.getLogger(__name__)
+console = Console()
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+EXPECTED_ATTRIBUTES = [
+    "name",
+    "type",
+    "description",
+    "logo",
+    "website",
+    "founded",
+    "ceo",
+    "headquarters",
+    "parent_organization",
+    "subsidiaries",
+    "social_twitter",
+    "social_facebook",
+    "social_linkedin",
+    "social_youtube",
+    "social_instagram",
+    "stock_ticker",
+    "industry",
+    "employees",
+    "revenue",
+]
+
+
+@dataclass
+class KnowledgePanelAttribute:
+    """Single attribute extracted from a Knowledge Panel."""
+    name: str
+    value: str | None = None
+    present: bool = False
+
+
+@dataclass
+class KnowledgePanel:
+    """Represents a detected Knowledge Panel."""
+    detected: bool = False
+    entity_type: str | None = None
+    attributes: list[KnowledgePanelAttribute] = field(default_factory=list)
+    completeness_score: float = 0.0
+    raw_snippet: str | None = None
+
+
+@dataclass
+class WikiPresence:
+    """Wikipedia or Wikidata presence record."""
+    platform: str = ""  # "wikipedia" or "wikidata"
+    present: bool = False
+    url: str | None = None
+    qid: str | None = None  # Wikidata QID (e.g. Q20710)
+    language: str = "en"
+
+
+@dataclass
+class NaverPresence:
+    """Naver encyclopedia and 지식iN presence."""
+    encyclopedia_present: bool = False
+    encyclopedia_url: str | None = None
+    knowledge_in_present: bool = False
+    knowledge_in_count: int = 0
+    knowledge_in_url: str | None = None
+
+
+@dataclass
+class KnowledgeGraphResult:
+    """Full Knowledge Graph analysis result."""
+    entity: str = ""
+    language: str = "en"
+    knowledge_panel: KnowledgePanel = field(default_factory=KnowledgePanel)
+    wikipedia: WikiPresence = field(default_factory=lambda: WikiPresence(platform="wikipedia"))
+    wikidata: WikiPresence = field(default_factory=lambda: WikiPresence(platform="wikidata"))
+    naver: NaverPresence = field(default_factory=NaverPresence)
+    competitors: list[dict[str, Any]] = field(default_factory=list)
+    overall_score: float = 0.0
+    recommendations: list[str] = field(default_factory=list)
+    timestamp: str = field(default_factory=lambda: datetime.now().isoformat())
+
+    def to_dict(self) -> dict[str, Any]:
+        return asdict(self)
+
+
+# ---------------------------------------------------------------------------
+# Knowledge Graph Analyzer
+# ---------------------------------------------------------------------------
+
+
+class KnowledgeGraphAnalyzer(BaseAsyncClient):
+    """Analyze entity presence in Knowledge Graph and related platforms."""
+
+    GOOGLE_SEARCH_URL = "https://www.google.com/search"
+    WIKIPEDIA_API_URL = "https://{lang}.wikipedia.org/api/rest_v1/page/summary/{title}"
+    WIKIDATA_API_URL = "https://www.wikidata.org/w/api.php"
+    NAVER_SEARCH_URL = "https://search.naver.com/search.naver"
+    NAVER_ENCYCLOPEDIA_URL = "https://terms.naver.com/search.naver"
+    NAVER_KIN_URL = "https://kin.naver.com/search/list.naver"
+
+    HEADERS = {
+        "User-Agent": (
+            "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
+            "AppleWebKit/537.36 (KHTML, like Gecko) "
+            "Chrome/120.0.0.0 Safari/537.36"
+        ),
+        "Accept-Language": "en-US,en;q=0.9",
+    }
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.config = config
+
+    # ------------------------------------------------------------------
+    # Google entity search
+    # ------------------------------------------------------------------
+
+    async def search_entity(
+        self,
+        entity_name: str,
+        language: str = "en",
+        session: aiohttp.ClientSession | None = None,
+    ) -> dict[str, Any]:
+        """Search Google for entity to detect Knowledge Panel signals."""
+        params = {"q": entity_name, "hl": language, "gl": "us" if language == "en" else "kr"}
+        headers = {**self.HEADERS}
+        if language == "ko":
+            headers["Accept-Language"] = "ko-KR,ko;q=0.9"
+            params["gl"] = "kr"
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            async with session.get(
+                self.GOOGLE_SEARCH_URL, params=params, headers=headers, timeout=aiohttp.ClientTimeout(total=20)
+            ) as resp:
+                if resp.status != 200:
+                    logger.warning("Google search returned status %d", resp.status)
+                    return {"html": "", "status": resp.status}
+                html = await resp.text()
+                return {"html": html, "status": resp.status}
+        except Exception as exc:
+            logger.error("Google search failed: %s", exc)
+            return {"html": "", "status": 0, "error": str(exc)}
+        finally:
+            if own_session:
+                await session.close()
+
+    # ------------------------------------------------------------------
+    # Knowledge Panel detection
+    # ------------------------------------------------------------------
+
+    def detect_knowledge_panel(self, search_data: dict[str, Any]) -> KnowledgePanel:
+        """Parse search results HTML for Knowledge Panel indicators."""
+        html = search_data.get("html", "")
+        if not html:
+            return KnowledgePanel(detected=False)
+
+        soup = BeautifulSoup(html, "lxml")
+        kp = KnowledgePanel()
+
+        # Knowledge Panel is typically in a div with class 'kp-wholepage' or 'knowledge-panel'
+        kp_selectors = [
+            "div.kp-wholepage",
+            "div.knowledge-panel",
+            "div[data-attrid='title']",
+            "div.kp-header",
+            "div[class*='kno-']",
+            "div.osrp-blk",
+        ]
+
+        kp_element = None
+        for selector in kp_selectors:
+            kp_element = soup.select_one(selector)
+            if kp_element:
+                break
+
+        if kp_element:
+            kp.detected = True
+            kp.raw_snippet = kp_element.get_text(separator=" ", strip=True)[:500]
+        else:
+            # Fallback: check for common KP text patterns
+            text = soup.get_text(separator=" ", strip=True).lower()
+            kp_indicators = [
+                "wikipedia", "description", "founded", "ceo",
+                "headquarters", "subsidiaries", "parent organization",
+            ]
+            matches = sum(1 for ind in kp_indicators if ind in text)
+            if matches >= 3:
+                kp.detected = True
+                kp.raw_snippet = text[:500]
+
+        return kp
+
+    # ------------------------------------------------------------------
+    # Attribute extraction
+    # ------------------------------------------------------------------
+
+    def extract_attributes(self, kp: KnowledgePanel, html: str = "") -> list[KnowledgePanelAttribute]:
+        """Extract entity attributes from Knowledge Panel data."""
+        attributes: list[KnowledgePanelAttribute] = []
+        text = (kp.raw_snippet or "").lower()
+
+        # Parse HTML for structured attribute data
+        soup = BeautifulSoup(html, "lxml") if html else None
+
+        attribute_patterns = {
+            "name": r"^(.+?)(?:\s+is\s+|\s*[-|]\s*)",
+            "type": r"(?:is\s+(?:a|an)\s+)(\w[\w\s]+?)(?:\.|,|\s+based)",
+            "description": r"(?:is\s+)(.{20,200}?)(?:\.\s)",
+            "founded": r"(?:founded|established|incorporated)\s*(?:in|:)?\s*(\d{4})",
+            "ceo": r"(?:ceo|chief executive|chairman)\s*(?::|is)?\s*([A-Z][\w\s.]+?)(?:,|\.|;|\s{2})",
+            "headquarters": r"(?:headquarters?|hq|based in)\s*(?::|is|in)?\s*([A-Z][\w\s,]+?)(?:\.|;|\s{2})",
+            "stock_ticker": r"(?:stock|ticker|symbol)\s*(?::|is)?\s*([A-Z]{1,5}(?:\s*:\s*[A-Z]{1,5})?)",
+            "employees": r"(?:employees?|staff|workforce)\s*(?::|is)?\s*([\d,]+)",
+            "revenue": r"(?:revenue|sales)\s*(?::|is)?\s*([\$\d,.]+\s*(?:billion|million|B|M)?)",
+            "industry": r"(?:industry|sector)\s*(?::|is)?\s*([\w\s&]+?)(?:\.|,|;)",
+        }
+
+        social_patterns = {
+            "social_twitter": r"(?:twitter\.com|x\.com)/(\w+)",
+            "social_facebook": r"facebook\.com/([\w.]+)",
+            "social_linkedin": r"linkedin\.com/(?:company|in)/([\w-]+)",
+            "social_youtube": r"youtube\.com/(?:@|channel/|user/)([\w-]+)",
+            "social_instagram": r"instagram\.com/([\w.]+)",
+        }
+
+        full_text = kp.raw_snippet or ""
+        html_text = ""
+        if soup:
+            html_text = soup.get_text(separator=" ", strip=True)
+
+        combined = f"{full_text} {html_text}"
+
+        for attr_name, pattern in attribute_patterns.items():
+            match = re.search(pattern, combined, re.IGNORECASE)
+            present = match is not None
+            value = match.group(1).strip() if match else None
+            attributes.append(KnowledgePanelAttribute(name=attr_name, value=value, present=present))
+
+        # Social profiles
+        for attr_name, pattern in social_patterns.items():
+            match = re.search(pattern, combined, re.IGNORECASE)
+            present = match is not None
+            value = match.group(1).strip() if match else None
+            attributes.append(KnowledgePanelAttribute(name=attr_name, value=value, present=present))
+
+        # Logo detection from HTML
+        logo_present = False
+        if soup:
+            logo_img = soup.select_one("img[data-atf], g-img img, img.kno-fb-img, img[alt*='logo']")
+            if logo_img:
+                logo_present = True
+        attributes.append(KnowledgePanelAttribute(name="logo", value=None, present=logo_present))
+
+        # Website detection
+        website_present = False
+        if soup:
+            site_link = soup.select_one("a[data-attrid*='website'], a.ab_button[href*='http']")
+            if site_link:
+                website_present = True
+                value = site_link.get("href", "")
+        attributes.append(KnowledgePanelAttribute(name="website", value=value if website_present else None, present=website_present))
+
+        return attributes
+
+    # ------------------------------------------------------------------
+    # Completeness scoring
+    # ------------------------------------------------------------------
+
+    def score_completeness(self, attributes: list[KnowledgePanelAttribute]) -> float:
+        """Score attribute completeness (0-100) based on filled vs expected."""
+        if not attributes:
+            return 0.0
+
+        weights = {
+            "name": 10, "type": 8, "description": 10, "logo": 8, "website": 10,
+            "founded": 5, "ceo": 5, "headquarters": 5, "parent_organization": 3,
+            "subsidiaries": 3, "social_twitter": 4, "social_facebook": 4,
+            "social_linkedin": 4, "social_youtube": 3, "social_instagram": 3,
+            "stock_ticker": 3, "industry": 5, "employees": 3, "revenue": 4,
+        }
+
+        total_weight = sum(weights.values())
+        earned = 0.0
+
+        attr_map = {a.name: a for a in attributes}
+        for attr_name, weight in weights.items():
+            attr = attr_map.get(attr_name)
+            if attr and attr.present:
+                earned += weight
+
+        return round((earned / total_weight) * 100, 1) if total_weight > 0 else 0.0
+
+    # ------------------------------------------------------------------
+    # Wikipedia check
+    # ------------------------------------------------------------------
+
+    async def check_wikipedia(
+        self,
+        entity_name: str,
+        language: str = "en",
+        session: aiohttp.ClientSession | None = None,
+    ) -> WikiPresence:
+        """Check Wikipedia article existence for entity."""
+        wiki = WikiPresence(platform="wikipedia", language=language)
+        title = entity_name.replace(" ", "_")
+        url = self.WIKIPEDIA_API_URL.format(lang=language, title=quote(title))
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            async with session.get(url, headers=self.HEADERS, timeout=aiohttp.ClientTimeout(total=15)) as resp:
+                if resp.status == 200:
+                    data = await resp.json()
+                    wiki.present = data.get("type") != "disambiguation"
+                    wiki.url = data.get("content_urls", {}).get("desktop", {}).get("page", "")
+                    if not wiki.url:
+                        wiki.url = f"https://{language}.wikipedia.org/wiki/{quote(title)}"
+                    logger.info("Wikipedia article found for '%s' (%s)", entity_name, language)
+                elif resp.status == 404:
+                    wiki.present = False
+                    logger.info("No Wikipedia article for '%s' (%s)", entity_name, language)
+                else:
+                    logger.warning("Wikipedia API returned status %d", resp.status)
+        except Exception as exc:
+            logger.error("Wikipedia check failed: %s", exc)
+        finally:
+            if own_session:
+                await session.close()
+
+        return wiki
+
+    # ------------------------------------------------------------------
+    # Wikidata check
+    # ------------------------------------------------------------------
+
+    async def check_wikidata(
+        self,
+        entity_name: str,
+        session: aiohttp.ClientSession | None = None,
+    ) -> WikiPresence:
+        """Check Wikidata QID existence for entity."""
+        wiki = WikiPresence(platform="wikidata")
+        params = {
+            "action": "wbsearchentities",
+            "search": entity_name,
+            "language": "en",
+            "format": "json",
+            "limit": 5,
+        }
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            async with session.get(
+                self.WIKIDATA_API_URL, params=params, headers=self.HEADERS,
+                timeout=aiohttp.ClientTimeout(total=15),
+            ) as resp:
+                if resp.status == 200:
+                    data = await resp.json()
+                    results = data.get("search", [])
+                    if results:
+                        top = results[0]
+                        wiki.present = True
+                        wiki.qid = top.get("id", "")
+                        wiki.url = top.get("concepturi", f"https://www.wikidata.org/wiki/{wiki.qid}")
+                        logger.info("Wikidata entity found: %s (%s)", wiki.qid, entity_name)
+                    else:
+                        wiki.present = False
+                        logger.info("No Wikidata entity for '%s'", entity_name)
+                else:
+                    logger.warning("Wikidata API returned status %d", resp.status)
+        except Exception as exc:
+            logger.error("Wikidata check failed: %s", exc)
+        finally:
+            if own_session:
+                await session.close()
+
+        return wiki
+
+    # ------------------------------------------------------------------
+    # Naver encyclopedia
+    # ------------------------------------------------------------------
+
+    async def check_naver_encyclopedia(
+        self,
+        entity_name: str,
+        session: aiohttp.ClientSession | None = None,
+    ) -> dict[str, Any]:
+        """Check Naver encyclopedia (네이버 백과사전) presence."""
+        result = {"present": False, "url": None}
+        params = {"query": entity_name, "searchType": 0}
+        headers = {
+            **self.HEADERS,
+            "Accept-Language": "ko-KR,ko;q=0.9",
+        }
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            async with session.get(
+                self.NAVER_ENCYCLOPEDIA_URL, params=params, headers=headers,
+                timeout=aiohttp.ClientTimeout(total=15),
+            ) as resp:
+                if resp.status == 200:
+                    html = await resp.text()
+                    soup = BeautifulSoup(html, "lxml")
+                    # Look for search result entries
+                    entries = soup.select("ul.content_list li, div.search_result a, a.title")
+                    if entries:
+                        result["present"] = True
+                        first_link = entries[0].find("a")
+                        if first_link and first_link.get("href"):
+                            href = first_link["href"]
+                            if not href.startswith("http"):
+                                href = urljoin("https://terms.naver.com", href)
+                            result["url"] = href
+                        else:
+                            result["url"] = f"https://terms.naver.com/search.naver?query={quote(entity_name)}"
+                        logger.info("Naver encyclopedia entry found for '%s'", entity_name)
+                    else:
+                        # Fallback: check page text for result indicators
+                        text = soup.get_text()
+                        if entity_name in text and "검색결과가 없습니다" not in text:
+                            result["present"] = True
+                            result["url"] = f"https://terms.naver.com/search.naver?query={quote(entity_name)}"
+                else:
+                    logger.warning("Naver encyclopedia returned status %d", resp.status)
+        except Exception as exc:
+            logger.error("Naver encyclopedia check failed: %s", exc)
+        finally:
+            if own_session:
+                await session.close()
+
+        return result
+
+    # ------------------------------------------------------------------
+    # Naver knowledge iN
+    # ------------------------------------------------------------------
+
+    async def check_naver_knowledge_in(
+        self,
+        entity_name: str,
+        session: aiohttp.ClientSession | None = None,
+    ) -> dict[str, Any]:
+        """Check Naver knowledge iN (지식iN) entries."""
+        result = {"present": False, "count": 0, "url": None}
+        params = {"query": entity_name}
+        headers = {
+            **self.HEADERS,
+            "Accept-Language": "ko-KR,ko;q=0.9",
+        }
+
+        own_session = session is None
+        if own_session:
+            session = aiohttp.ClientSession()
+
+        try:
+            async with session.get(
+                self.NAVER_KIN_URL, params=params, headers=headers,
+                timeout=aiohttp.ClientTimeout(total=15),
+            ) as resp:
+                if resp.status == 200:
+                    html = await resp.text()
+                    soup = BeautifulSoup(html, "lxml")
+
+                    # Extract total result count
+                    count_el = soup.select_one("span.number, em.total_count, span.result_count")
+                    count = 0
+                    if count_el:
+                        count_text = count_el.get_text(strip=True).replace(",", "")
+                        count_match = re.search(r"(\d+)", count_text)
+                        if count_match:
+                            count = int(count_match.group(1))
+
+                    # Also check for list items
+                    entries = soup.select("ul.basic1 li, ul._list li, div.search_list li")
+                    if count > 0 or entries:
+                        result["present"] = True
+                        result["count"] = count if count > 0 else len(entries)
+                        result["url"] = f"https://kin.naver.com/search/list.naver?query={quote(entity_name)}"
+                        logger.info("Naver 지식iN: %d entries for '%s'", result["count"], entity_name)
+                    else:
+                        logger.info("No Naver 지식iN entries for '%s'", entity_name)
+                else:
+                    logger.warning("Naver 지식iN returned status %d", resp.status)
+        except Exception as exc:
+            logger.error("Naver 지식iN check failed: %s", exc)
+        finally:
+            if own_session:
+                await session.close()
+
+        return result
+
+    # ------------------------------------------------------------------
+    # Recommendations
+    # ------------------------------------------------------------------
+
+    def generate_recommendations(self, result: KnowledgeGraphResult) -> list[str]:
+        """Generate actionable recommendations based on analysis."""
+        recs: list[str] = []
+
+        kp = result.knowledge_panel
+        if not kp.detected:
+            recs.append(
+                "Knowledge Panel이 감지되지 않았습니다. Google에 엔티티 등록을 위해 "
+                "Wikipedia 페이지 생성, Wikidata 항목 추가, 구조화된 데이터(Organization schema) 구현을 권장합니다."
+            )
+        elif kp.completeness_score < 50:
+            recs.append(
+                f"Knowledge Panel 완성도가 {kp.completeness_score}%로 낮습니다. "
+                "누락된 속성(소셜 프로필, 설명, 로고 등)을 보강하세요."
+            )
+
+        if not result.wikipedia.present:
+            recs.append(
+                "Wikipedia 문서가 없습니다. 주목할 만한 출처(reliable sources)를 확보한 후 "
+                "Wikipedia 문서 생성을 고려하세요."
+            )
+
+        if not result.wikidata.present:
+            recs.append(
+                "Wikidata 항목이 없습니다. Wikidata에 엔티티를 등록하여 "
+                "Knowledge Graph 인식을 강화하세요."
+            )
+
+        if not result.naver.encyclopedia_present:
+            recs.append(
+                "네이버 백과사전에 등록되어 있지 않습니다. 한국 시장 SEO를 위해 "
+                "네이버 백과사전 등재를 검토하세요."
+            )
+
+        if result.naver.knowledge_in_count < 5:
+            recs.append(
+                "네이버 지식iN에 관련 콘텐츠가 부족합니다. Q&A 콘텐츠를 통해 "
+                "브랜드 엔티티 인지도를 높이세요."
+            )
+
+        # Check social profile completeness
+        attr_map = {a.name: a for a in kp.attributes}
+        missing_social = []
+        for soc in ["social_twitter", "social_facebook", "social_linkedin", "social_youtube"]:
+            attr = attr_map.get(soc)
+            if not attr or not attr.present:
+                missing_social.append(soc.replace("social_", "").title())
+        if missing_social:
+            recs.append(
+                f"소셜 프로필 연결 누락: {', '.join(missing_social)}. "
+                "웹사이트 schema의 sameAs 속성에 소셜 프로필을 추가하세요."
+            )
+
+        if not recs:
+            recs.append("Knowledge Graph 엔티티 상태가 양호합니다. 현재 수준을 유지하세요.")
+
+        return recs
+
+    # ------------------------------------------------------------------
+    # Main orchestrator
+    # ------------------------------------------------------------------
+
+    async def analyze(
+        self,
+        entity_name: str,
+        language: str = "en",
+        include_wiki: bool = True,
+        include_naver: bool = True,
+    ) -> KnowledgeGraphResult:
+        """Orchestrate full Knowledge Graph analysis."""
+        result = KnowledgeGraphResult(entity=entity_name, language=language)
+        logger.info("Starting Knowledge Graph analysis for '%s' (lang=%s)", entity_name, language)
+
+        async with aiohttp.ClientSession() as session:
+            # Step 1: Search entity on Google
+            search_data = await self.search_entity(entity_name, language, session)
+
+            # Step 2: Detect Knowledge Panel
+            kp = self.detect_knowledge_panel(search_data)
+
+            # Step 3: Extract attributes
+            if kp.detected:
+                kp.attributes = self.extract_attributes(kp, search_data.get("html", ""))
+                kp.completeness_score = self.score_completeness(kp.attributes)
+
+                # Detect entity type from attributes
+                for attr in kp.attributes:
+                    if attr.name == "type" and attr.present:
+                        kp.entity_type = attr.value
+                        break
+
+            result.knowledge_panel = kp
+
+            # Step 4: Wikipedia and Wikidata checks (parallel)
+            if include_wiki:
+                wiki_task = self.check_wikipedia(entity_name, language, session)
+                wikidata_task = self.check_wikidata(entity_name, session)
+                result.wikipedia, result.wikidata = await asyncio.gather(wiki_task, wikidata_task)
+
+            # Step 5: Naver checks (parallel)
+            if include_naver:
+                enc_task = self.check_naver_encyclopedia(entity_name, session)
+                kin_task = self.check_naver_knowledge_in(entity_name, session)
+                enc_result, kin_result = await asyncio.gather(enc_task, kin_task)
+
+                result.naver = NaverPresence(
+                    encyclopedia_present=enc_result.get("present", False),
+                    encyclopedia_url=enc_result.get("url"),
+                    knowledge_in_present=kin_result.get("present", False),
+                    knowledge_in_count=kin_result.get("count", 0),
+                    knowledge_in_url=kin_result.get("url"),
+                )
+
+        # Step 6: Compute overall score
+        scores = []
+        if kp.detected:
+            scores.append(kp.completeness_score * 0.35)
+        else:
+            scores.append(0)
+        scores.append(20.0 if result.wikipedia.present else 0)
+        scores.append(15.0 if result.wikidata.present else 0)
+        scores.append(15.0 if result.naver.encyclopedia_present else 0)
+        scores.append(15.0 if result.naver.knowledge_in_present else 0)
+        result.overall_score = round(sum(scores), 1)
+
+        # Step 7: Recommendations
+        result.recommendations = self.generate_recommendations(result)
+
+        logger.info("Analysis complete. Overall score: %.1f", result.overall_score)
+        return result
+
+
+# ---------------------------------------------------------------------------
+# CLI display helpers
+# ---------------------------------------------------------------------------
+
+
+def display_result(result: KnowledgeGraphResult) -> None:
+    """Display analysis result in a rich table."""
+    console.print()
+    console.print(f"[bold cyan]Knowledge Graph Analysis: {result.entity}[/bold cyan]")
+    console.print(f"Language: {result.language} | Score: {result.overall_score}/100")
+    console.print()
+
+    # Knowledge Panel table
+    kp = result.knowledge_panel
+    table = Table(title="Knowledge Panel", show_header=True)
+    table.add_column("Property", style="bold")
+    table.add_column("Value")
+    table.add_column("Status")
+
+    table.add_row("Detected", str(kp.detected), "[green]OK[/]" if kp.detected else "[red]Missing[/]")
+    table.add_row("Entity Type", kp.entity_type or "-", "[green]OK[/]" if kp.entity_type else "[yellow]Unknown[/]")
+    table.add_row("Completeness", f"{kp.completeness_score}%", "[green]OK[/]" if kp.completeness_score >= 50 else "[red]Low[/]")
+
+    for attr in kp.attributes:
+        status = "[green]Present[/]" if attr.present else "[red]Missing[/]"
+        table.add_row(f"  {attr.name}", attr.value or "-", status)
+
+    console.print(table)
+    console.print()
+
+    # Platform presence table
+    plat_table = Table(title="Platform Presence", show_header=True)
+    plat_table.add_column("Platform", style="bold")
+    plat_table.add_column("Present")
+    plat_table.add_column("Details")
+
+    plat_table.add_row(
+        "Wikipedia",
+        "[green]Yes[/]" if result.wikipedia.present else "[red]No[/]",
+        result.wikipedia.url or "-",
+    )
+    plat_table.add_row(
+        "Wikidata",
+        "[green]Yes[/]" if result.wikidata.present else "[red]No[/]",
+        result.wikidata.qid or "-",
+    )
+    plat_table.add_row(
+        "Naver Encyclopedia",
+        "[green]Yes[/]" if result.naver.encyclopedia_present else "[red]No[/]",
+        result.naver.encyclopedia_url or "-",
+    )
+    plat_table.add_row(
+        "Naver 지식iN",
+        "[green]Yes[/]" if result.naver.knowledge_in_present else "[red]No[/]",
+        f"{result.naver.knowledge_in_count} entries" if result.naver.knowledge_in_present else "-",
+    )
+
+    console.print(plat_table)
+    console.print()
+
+    # Recommendations
+    console.print("[bold yellow]Recommendations:[/bold yellow]")
+    for i, rec in enumerate(result.recommendations, 1):
+        console.print(f"  {i}. {rec}")
+    console.print()
+
+
+# ---------------------------------------------------------------------------
+# Main
+# ---------------------------------------------------------------------------
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Knowledge Graph & Entity Analyzer",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    parser.add_argument("--entity", required=True, help="Entity name to analyze")
+    parser.add_argument("--language", default="en", choices=["en", "ko", "ja", "zh"], help="Language (default: en)")
+    parser.add_argument("--wiki", action="store_true", default=True, help="Include Wikipedia/Wikidata check (default: True)")
+    parser.add_argument("--no-wiki", action="store_true", help="Skip Wikipedia/Wikidata check")
+    parser.add_argument("--no-naver", action="store_true", help="Skip Naver checks")
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+    parser.add_argument("--output", type=str, help="Output file path")
+    return parser.parse_args()
+
+
+async def main() -> None:
+    args = parse_args()
+
+    analyzer = KnowledgeGraphAnalyzer()
+    result = await analyzer.analyze(
+        entity_name=args.entity,
+        language=args.language,
+        include_wiki=not args.no_wiki,
+        include_naver=not args.no_naver,
+    )
+
+    if args.json:
+        output = json.dumps(result.to_dict(), ensure_ascii=False, indent=2)
+        if args.output:
+            with open(args.output, "w", encoding="utf-8") as f:
+                f.write(output)
+            console.print(f"[green]Output saved to {args.output}[/green]")
+        else:
+            print(output)
+    else:
+        display_result(result)
+        if args.output:
+            with open(args.output, "w", encoding="utf-8") as f:
+                json.dump(result.to_dict(), f, ensure_ascii=False, indent=2)
+            console.print(f"[green]Output saved to {args.output}[/green]")
+
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/custom-skills/28-seo-knowledge-graph/code/scripts/requirements.txt b/custom-skills/28-seo-knowledge-graph/code/scripts/requirements.txt
new file mode 100644
index 0000000..5cb1da4
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/code/scripts/requirements.txt
@@ -0,0 +1,9 @@
+# 28-seo-knowledge-graph dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+beautifulsoup4>=4.12.0
+lxml>=5.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/28-seo-knowledge-graph/desktop/SKILL.md b/custom-skills/28-seo-knowledge-graph/desktop/SKILL.md
new file mode 100644
index 0000000..37f75ce
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/desktop/SKILL.md
@@ -0,0 +1,77 @@
+---
+name: seo-knowledge-graph
+description: |
+  Knowledge Graph and entity SEO analysis. Triggers: knowledge panel, entity SEO, knowledge graph, PAA, FAQ schema, Wikipedia, Wikidata, brand entity.
+---
+
+# Knowledge Graph & Entity SEO
+
+Analyze brand entity presence in Google Knowledge Graph, Knowledge Panels, People Also Ask (PAA), and FAQ rich results. Check entity attribute completeness, Wikipedia/Wikidata presence, and Korean equivalents (Naver knowledge iN, Naver encyclopedia).
+
+## Capabilities
+
+### Knowledge Graph Analysis
+- Knowledge Panel detection and attribute extraction
+- Entity attribute completeness scoring (name, description, logo, type, social profiles, website, founded, CEO)
+- Wikipedia article presence check
+- Wikidata entity presence check (QID lookup)
+- Naver encyclopedia (네이버 백과사전) presence
+- Naver knowledge iN (지식iN) presence
+
+### Entity SEO Audit
+- People Also Ask (PAA) monitoring for brand-related queries
+- FAQ schema presence tracking (FAQPage schema -> SERP appearance)
+- Entity markup audit (Organization, Person, LocalBusiness schema on website)
+- Social profile linking validation (sameAs in schema)
+- Brand SERP analysis (what appears when you search the brand name)
+- Entity consistency across web properties
+
+## Workflow
+
+### Knowledge Graph Analysis
+
+1. Use **WebSearch** to search for the entity name on Google
+2. Analyze search results for Knowledge Panel indicators
+3. Use **WebFetch** to check Wikipedia article existence
+4. Use **WebFetch** to check Wikidata QID existence
+5. Use **WebFetch** to check Naver encyclopedia and 지식iN
+6. Score entity attribute completeness
+7. Save report to **Notion** SEO Audit Log
+
+### Entity SEO Audit
+
+1. Use **WebFetch** to fetch the website and extract JSON-LD schemas
+2. Validate Organization/Person/LocalBusiness schema completeness
+3. Check sameAs links accessibility
+4. Use **WebSearch** to search brand name and analyze SERP features
+5. Monitor PAA questions for brand keywords
+6. Use **Ahrefs serp-overview** for SERP feature detection
+7. Save report to **Notion** SEO Audit Log
+
+## Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| WebSearch | Search for entity/brand to detect Knowledge Panel |
+| WebFetch | Fetch Wikipedia, Wikidata, Naver pages, website schemas |
+| Ahrefs `serp-overview` | SERP feature detection for entity keywords |
+| Notion | Save audit reports to SEO Audit Log database |
+
+## Notion Output
+
+All reports must be saved to the OurDigital SEO Audit Log database.
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| Category | Knowledge Graph & Entity SEO |
+| Audit ID | KG-YYYYMMDD-NNN |
+
+Report content should be written in Korean (한국어), keeping technical English terms as-is.
+
+## Reference Scripts
+
+Located in `code/scripts/`:
+- `knowledge_graph_analyzer.py` -- Knowledge Panel and entity presence analysis
+- `entity_auditor.py` -- Entity SEO signals and PAA/FAQ audit
+- `base_client.py` -- Shared async client utilities
diff --git a/custom-skills/28-seo-knowledge-graph/desktop/skill.yaml b/custom-skills/28-seo-knowledge-graph/desktop/skill.yaml
new file mode 100644
index 0000000..ded87ca
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-knowledge-graph
+description: |
+  Knowledge Graph and entity SEO analysis. Triggers: knowledge panel, entity SEO, knowledge graph, PAA, FAQ schema, Wikipedia, Wikidata, brand entity.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/28-seo-knowledge-graph/desktop/tools/ahrefs.md b/custom-skills/28-seo-knowledge-graph/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..0abb2d1
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/desktop/tools/ahrefs.md
@@ -0,0 +1,29 @@
+# Ahrefs MCP Tools
+
+## serp-overview
+
+Retrieve SERP feature overview for specified keywords. Used to detect Knowledge Panel, PAA, FAQ rich results, and other SERP features for entity-related queries.
+
+### Parameters
+
+| Parameter | Type | Required | Description |
+|-----------|------|----------|-------------|
+| keyword | string | Yes | Keyword to check SERP features for |
+| country | string | No | Country code (default: us) |
+
+### Usage
+
+```
+mcp__ahrefs__serp-overview(keyword="Samsung Electronics", country="us")
+```
+
+### Relevant SERP Features
+
+- Knowledge Panel
+- People Also Ask (PAA)
+- FAQ Rich Results
+- Featured Snippet
+- Sitelinks
+- Image Pack
+- Video Carousel
+- Local Pack
diff --git a/custom-skills/28-seo-knowledge-graph/desktop/tools/notion.md b/custom-skills/28-seo-knowledge-graph/desktop/tools/notion.md
new file mode 100644
index 0000000..9062532
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/desktop/tools/notion.md
@@ -0,0 +1,54 @@
+# Notion MCP Tools
+
+## Save Audit Report
+
+Save Knowledge Graph & Entity SEO audit reports to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Entity website URL |
+| Category | Select | Knowledge Graph & Entity SEO |
+| Priority | Select | Critical / High / Medium / Low |
+| Found Date | Date | Audit date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: KG-YYYYMMDD-NNN |
+
+### Priority Mapping
+
+| Overall Score | Priority |
+|---------------|----------|
+| 0-25 | Critical |
+| 26-50 | High |
+| 51-75 | Medium |
+| 76-100 | Low |
+
+### Example Usage
+
+```
+mcp__notion__notion-create-pages({
+  "parent": {"database_id": "2c8581e5-8a1e-8035-880b-e38cefc2f3ef"},
+  "properties": {
+    "Issue": {"title": [{"text": {"content": "Knowledge Graph 엔티티 분석 - 2025-01-15"}}]},
+    "Site": {"url": "https://example.com"},
+    "Category": {"select": {"name": "Knowledge Graph & Entity SEO"}},
+    "Priority": {"select": {"name": "High"}},
+    "Found Date": {"date": {"start": "2025-01-15"}},
+    "Audit ID": {"rich_text": [{"text": {"content": "KG-20250115-001"}}]}
+  }
+})
+```
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (Knowledge Panel, Knowledge Graph, PAA, FAQ, schema)
+- URLs and code remain unchanged
diff --git a/custom-skills/28-seo-knowledge-graph/desktop/tools/websearch.md b/custom-skills/28-seo-knowledge-graph/desktop/tools/websearch.md
new file mode 100644
index 0000000..82b2b28
--- /dev/null
+++ b/custom-skills/28-seo-knowledge-graph/desktop/tools/websearch.md
@@ -0,0 +1,62 @@
+# WebSearch & WebFetch Tools
+
+## WebSearch
+
+Search the web for entity/brand information to detect Knowledge Panel signals, PAA questions, and SERP features.
+
+### Usage Patterns
+
+#### Knowledge Panel Detection
+```
+WebSearch("Samsung Electronics")
+WebSearch("삼성전자")
+```
+
+#### PAA Question Discovery
+```
+WebSearch("what is Samsung Electronics")
+WebSearch("Samsung Electronics reviews")
+WebSearch("Samsung Electronics alternatives")
+```
+
+#### Brand SERP Analysis
+```
+WebSearch("Brand Name official")
+WebSearch("Brand Name site:wikipedia.org")
+```
+
+## WebFetch
+
+Fetch and analyze specific pages for entity presence data.
+
+### Usage Patterns
+
+#### Wikipedia Check
+```
+WebFetch("https://en.wikipedia.org/wiki/Samsung_Electronics", "Check if article exists and extract entity information")
+```
+
+#### Wikidata Check
+```
+WebFetch("https://www.wikidata.org/w/api.php?action=wbsearchentities&search=Samsung&language=en&format=json", "Find Wikidata QID for entity")
+```
+
+#### Naver Encyclopedia
+```
+WebFetch("https://terms.naver.com/search.naver?query=삼성전자", "Check Naver encyclopedia entry")
+```
+
+#### Naver Knowledge iN
+```
+WebFetch("https://kin.naver.com/search/list.naver?query=삼성전자", "Check knowledge iN entries count")
+```
+
+#### Website Schema Extraction
+```
+WebFetch("https://example.com", "Extract JSON-LD structured data, especially Organization, Person, LocalBusiness schemas and sameAs links")
+```
+
+#### FAQ Schema Check
+```
+WebFetch("https://example.com/faq", "Check for FAQPage JSON-LD schema markup")
+```
diff --git a/custom-skills/31-seo-competitor-intel/code/CLAUDE.md b/custom-skills/31-seo-competitor-intel/code/CLAUDE.md
new file mode 100644
index 0000000..fe617d6
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/code/CLAUDE.md
@@ -0,0 +1,153 @@
+# CLAUDE.md
+
+## Overview
+
+Competitor intelligence and benchmarking tool for comprehensive SEO competitive analysis. Auto-discovers competitors via Ahrefs, builds competitor profile cards (DR, traffic, keywords, backlinks, content volume), creates head-to-head comparison matrices, tracks traffic trends, analyzes keyword overlap, compares content freshness/volume, and scores competitive threats. Supports Korean market competitor analysis including Naver Blog/Cafe presence.
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Competitor profiling
+python scripts/competitor_profiler.py --target https://example.com --json
+
+# Competitive monitoring
+python scripts/competitive_monitor.py --target https://example.com --period 30 --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `competitor_profiler.py` | Build competitor profiles and comparison matrix | Competitor cards, head-to-head matrix, threat scores |
+| `competitive_monitor.py` | Track competitive changes over time | Traffic trends, keyword movement, content velocity |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## Competitor Profiler
+
+```bash
+# Auto-discover and profile competitors
+python scripts/competitor_profiler.py --target https://example.com --json
+
+# Specify competitors manually
+python scripts/competitor_profiler.py --target https://example.com --competitor https://comp1.com --competitor https://comp2.com --json
+
+# Limit auto-discovery count
+python scripts/competitor_profiler.py --target https://example.com --max-competitors 10 --json
+
+# Include Korean market analysis
+python scripts/competitor_profiler.py --target https://example.com --korean-market --json
+```
+
+**Capabilities**:
+- Competitor auto-discovery via Ahrefs organic-competitors
+- Competitor profile cards:
+  - Domain Rating (DR)
+  - Organic traffic estimate
+  - Total organic keywords
+  - Referring domains count
+  - Top pages count
+  - Content volume estimate
+- Head-to-head comparison matrix across all SEO dimensions
+- Keyword overlap analysis (shared, unique to target, unique to competitor, gap)
+- Content volume and freshness comparison
+- Competitive threat scoring (0-100) based on growth trajectory, keyword overlap, DR gap
+- Korean market: Naver Blog/Cafe presence detection for competitors
+
+## Competitive Monitor
+
+```bash
+# 30-day competitive monitoring
+python scripts/competitive_monitor.py --target https://example.com --period 30 --json
+
+# With specific competitors
+python scripts/competitive_monitor.py --target https://example.com --competitor https://comp1.com --period 60 --json
+
+# Traffic trend comparison
+python scripts/competitive_monitor.py --target https://example.com --scope traffic --period 90 --json
+```
+
+**Capabilities**:
+- Traffic trend comparison (6-12 month history)
+- DR trend comparison over time
+- New/lost keyword tracking per competitor
+- Referring domain growth comparison
+- Content publication velocity comparison
+- Top page changes (new top pages, declining pages)
+- Alert generation for significant competitive movements
+- Market share estimation based on organic traffic
+
+## Ahrefs MCP Tools Used
+
+| Tool | Purpose |
+|------|---------|
+| `site-explorer-metrics` | Current organic metrics |
+| `site-explorer-metrics-history` | Historical metrics trends |
+| `site-explorer-organic-competitors` | Discover organic competitors |
+| `site-explorer-organic-keywords` | Keyword rankings per domain |
+| `site-explorer-top-pages` | Top performing pages |
+| `site-explorer-pages-by-traffic` | Pages ranked by traffic |
+| `site-explorer-domain-rating` | Domain Rating |
+| `site-explorer-domain-rating-history` | DR trend |
+| `site-explorer-referring-domains` | Referring domain list |
+| `site-explorer-backlinks-stats` | Backlink overview |
+| `site-explorer-pages-history` | Page index history |
+
+## Output Format
+
+```json
+{
+  "target": "example.com",
+  "competitors": [
+    {
+      "domain": "competitor1.com",
+      "profile": {
+        "domain_rating": 55,
+        "organic_traffic": 25000,
+        "organic_keywords": 3500,
+        "referring_domains": 1200,
+        "top_pages": 150
+      },
+      "threat_score": 78,
+      "keyword_overlap": {
+        "shared": 450,
+        "unique_competitor": 800,
+        "unique_target": 600,
+        "gap": 350
+      }
+    }
+  ],
+  "comparison_matrix": {},
+  "market_position": "challenger",
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Target website URL |
+| Category | Select | Competitor Intelligence |
+| Priority | Select | Based on threat level |
+| Found Date | Date | Report date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: COMP-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., Domain Rating, Organic Traffic, Threat Score)
+- URLs and code remain unchanged
diff --git a/custom-skills/31-seo-competitor-intel/code/scripts/base_client.py b/custom-skills/31-seo-competitor-intel/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/31-seo-competitor-intel/code/scripts/competitive_monitor.py b/custom-skills/31-seo-competitor-intel/code/scripts/competitive_monitor.py
new file mode 100644
index 0000000..87f488d
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/code/scripts/competitive_monitor.py
@@ -0,0 +1,776 @@
+"""
+Competitive Monitor - Track SEO Competitive Changes Over Time
+=============================================================
+Purpose: Monitor traffic trends, DR changes, keyword movement,
+         content velocity, and generate competitive alerts.
+Python: 3.10+
+
+Usage:
+    python competitive_monitor.py --target https://example.com --period 30 --json
+    python competitive_monitor.py --target https://example.com --competitor https://comp1.com --period 60 --json
+    python competitive_monitor.py --target https://example.com --scope traffic --period 90 --json
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime, timedelta
+from typing import Any
+from urllib.parse import urlparse
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class TrafficTrend:
+    """Traffic trend for a domain over a time period."""
+    domain: str = ""
+    data_points: list[dict[str, Any]] = field(default_factory=list)
+    direction: str = "stable"  # up / down / stable
+    growth_rate: float = 0.0
+    current_traffic: int = 0
+    period_start_traffic: int = 0
+
+
+@dataclass
+class DrTrend:
+    """Domain Rating trend for a domain."""
+    domain: str = ""
+    data_points: list[dict[str, Any]] = field(default_factory=list)
+    direction: str = "stable"
+    current_dr: float = 0.0
+    period_start_dr: float = 0.0
+    change: float = 0.0
+
+
+@dataclass
+class KeywordMovement:
+    """Keyword gains and losses for a domain."""
+    domain: str = ""
+    new_keywords: int = 0
+    lost_keywords: int = 0
+    net_change: int = 0
+    improved_positions: int = 0
+    declined_positions: int = 0
+
+
+@dataclass
+class ContentVelocity:
+    """Content publication rate metrics."""
+    domain: str = ""
+    new_pages_per_month: float = 0.0
+    total_pages_start: int = 0
+    total_pages_end: int = 0
+    net_new_pages: int = 0
+    avg_word_count: int = 0
+
+
+@dataclass
+class CompetitiveAlert:
+    """Alert for significant competitive movement."""
+    domain: str = ""
+    alert_type: str = ""  # traffic_surge, dr_jump, keyword_surge, content_burst
+    message: str = ""
+    severity: str = "info"  # info / warning / critical
+    metric_value: float = 0.0
+    threshold: float = 0.0
+
+
+@dataclass
+class MarketShare:
+    """Market share estimation for a domain within its competitive set."""
+    domain: str = ""
+    traffic_share_pct: float = 0.0
+    keyword_share_pct: float = 0.0
+    overall_share_pct: float = 0.0
+
+
+@dataclass
+class CompetitiveMonitorResult:
+    """Full monitoring result."""
+    target: str = ""
+    period_days: int = 30
+    scope: str = "all"
+    traffic_trends: list[TrafficTrend] = field(default_factory=list)
+    dr_trends: list[DrTrend] = field(default_factory=list)
+    keyword_movements: list[KeywordMovement] = field(default_factory=list)
+    content_velocities: list[ContentVelocity] = field(default_factory=list)
+    alerts: list[CompetitiveAlert] = field(default_factory=list)
+    market_shares: list[MarketShare] = field(default_factory=list)
+    timestamp: str = ""
+    errors: list[str] = field(default_factory=list)
+
+
+# ---------------------------------------------------------------------------
+# Monitor
+# ---------------------------------------------------------------------------
+
+class CompetitiveMonitor(BaseAsyncClient):
+    """Monitor competitive landscape changes using Ahrefs MCP tools."""
+
+    # Alert thresholds
+    TRAFFIC_CHANGE_THRESHOLD = 0.20   # 20% change triggers alert
+    DR_CHANGE_THRESHOLD = 3.0         # 3-point DR change
+    KEYWORD_SURGE_THRESHOLD = 0.15    # 15% keyword growth
+    CONTENT_BURST_THRESHOLD = 2.0     # 2x normal content velocity
+
+    def __init__(self):
+        super().__init__(max_concurrent=5, requests_per_second=2.0)
+
+    @staticmethod
+    def _extract_domain(url: str) -> str:
+        """Extract bare domain from URL or return as-is if already bare."""
+        if "://" in url:
+            parsed = urlparse(url)
+            return parsed.netloc.lower().replace("www.", "")
+        return url.lower().replace("www.", "")
+
+    async def _call_ahrefs(self, tool: str, params: dict[str, Any]) -> dict:
+        """Simulate Ahrefs MCP call. In production, routed via MCP bridge."""
+        self.logger.info(f"Ahrefs MCP call: {tool} | params={params}")
+        return {"tool": tool, "params": params, "data": {}}
+
+    async def _discover_competitors(
+        self, target: str, limit: int = 10
+    ) -> list[str]:
+        """Discover competitors if none specified."""
+        domain = self._extract_domain(target)
+        resp = await self._call_ahrefs(
+            "site-explorer-organic-competitors",
+            {"target": domain, "limit": limit},
+        )
+        competitors_raw = resp.get("data", {}).get("competitors", [])
+        return [
+            c.get("domain", "")
+            for c in competitors_raw
+            if c.get("domain") and c.get("domain") != domain
+        ][:limit]
+
+    async def get_traffic_trends(
+        self, domains: list[str], period: int
+    ) -> list[TrafficTrend]:
+        """Get traffic trend data for multiple domains."""
+        trends: list[TrafficTrend] = []
+        date_from = (datetime.now() - timedelta(days=period)).strftime("%Y-%m-%d")
+
+        for domain in domains:
+            domain = self._extract_domain(domain)
+            trend = TrafficTrend(domain=domain)
+
+            resp = await self._call_ahrefs(
+                "site-explorer-metrics-history",
+                {"target": domain, "date_from": date_from},
+            )
+            data_points = resp.get("data", {}).get("data_points", [])
+            trend.data_points = data_points
+
+            if len(data_points) >= 2:
+                first_traffic = int(data_points[0].get("organic_traffic", 0))
+                last_traffic = int(data_points[-1].get("organic_traffic", 0))
+                trend.period_start_traffic = first_traffic
+                trend.current_traffic = last_traffic
+
+                if first_traffic > 0:
+                    growth = ((last_traffic - first_traffic) / first_traffic) * 100
+                    trend.growth_rate = round(growth, 2)
+
+                    if growth > 5:
+                        trend.direction = "up"
+                    elif growth < -5:
+                        trend.direction = "down"
+                    else:
+                        trend.direction = "stable"
+                else:
+                    trend.direction = "new" if last_traffic > 0 else "stable"
+
+            trends.append(trend)
+            self.logger.info(
+                f"Traffic trend for {domain}: {trend.direction} "
+                f"({trend.growth_rate:+.1f}%)"
+            )
+
+        return trends
+
+    async def get_dr_trends(
+        self, domains: list[str], period: int
+    ) -> list[DrTrend]:
+        """Get Domain Rating trend data for multiple domains."""
+        trends: list[DrTrend] = []
+        date_from = (datetime.now() - timedelta(days=period)).strftime("%Y-%m-%d")
+
+        for domain in domains:
+            domain = self._extract_domain(domain)
+            trend = DrTrend(domain=domain)
+
+            resp = await self._call_ahrefs(
+                "site-explorer-domain-rating-history",
+                {"target": domain, "date_from": date_from},
+            )
+            data_points = resp.get("data", {}).get("data_points", [])
+            trend.data_points = data_points
+
+            if len(data_points) >= 2:
+                first_dr = float(data_points[0].get("domain_rating", 0))
+                last_dr = float(data_points[-1].get("domain_rating", 0))
+                trend.period_start_dr = first_dr
+                trend.current_dr = last_dr
+                trend.change = round(last_dr - first_dr, 1)
+
+                if trend.change > 1:
+                    trend.direction = "up"
+                elif trend.change < -1:
+                    trend.direction = "down"
+                else:
+                    trend.direction = "stable"
+
+            trends.append(trend)
+            self.logger.info(
+                f"DR trend for {domain}: {trend.direction} "
+                f"(change={trend.change:+.1f})"
+            )
+
+        return trends
+
+    async def track_keyword_movement(
+        self, domains: list[str], period: int
+    ) -> list[KeywordMovement]:
+        """Track new/lost keywords for each domain over the period."""
+        movements: list[KeywordMovement] = []
+
+        for domain in domains:
+            domain = self._extract_domain(domain)
+            movement = KeywordMovement(domain=domain)
+
+            # Current keyword count
+            current_resp = await self._call_ahrefs(
+                "site-explorer-metrics",
+                {"target": domain},
+            )
+            current_kw = int(
+                current_resp.get("data", {}).get("organic_keywords", 0)
+            )
+
+            # Historical keyword count
+            date_from = (datetime.now() - timedelta(days=period)).strftime("%Y-%m-%d")
+            hist_resp = await self._call_ahrefs(
+                "site-explorer-metrics-history",
+                {"target": domain, "date_from": date_from},
+            )
+            data_points = hist_resp.get("data", {}).get("data_points", [])
+
+            if data_points:
+                start_kw = int(data_points[0].get("organic_keywords", 0))
+            else:
+                start_kw = current_kw
+
+            net_change = current_kw - start_kw
+            movement.net_change = net_change
+
+            # Estimate new vs lost (simplified: positive net = new > lost)
+            if net_change > 0:
+                movement.new_keywords = net_change
+                movement.lost_keywords = 0
+                movement.improved_positions = int(net_change * 0.6)
+                movement.declined_positions = int(net_change * 0.1)
+            elif net_change < 0:
+                movement.new_keywords = 0
+                movement.lost_keywords = abs(net_change)
+                movement.improved_positions = 0
+                movement.declined_positions = abs(net_change)
+            else:
+                movement.new_keywords = 0
+                movement.lost_keywords = 0
+
+            movements.append(movement)
+            self.logger.info(
+                f"Keyword movement for {domain}: net={movement.net_change:+d}"
+            )
+
+        return movements
+
+    async def compare_content_velocity(
+        self, domains: list[str], period: int
+    ) -> list[ContentVelocity]:
+        """Compare content publication velocity across domains."""
+        velocities: list[ContentVelocity] = []
+        date_from = (datetime.now() - timedelta(days=period)).strftime("%Y-%m-%d")
+
+        for domain in domains:
+            domain = self._extract_domain(domain)
+            velocity = ContentVelocity(domain=domain)
+
+            resp = await self._call_ahrefs(
+                "site-explorer-pages-history",
+                {"target": domain, "date_from": date_from},
+            )
+            data_points = resp.get("data", {}).get("data_points", [])
+
+            if len(data_points) >= 2:
+                start_pages = int(data_points[0].get("pages", 0))
+                end_pages = int(data_points[-1].get("pages", 0))
+                velocity.total_pages_start = start_pages
+                velocity.total_pages_end = end_pages
+                velocity.net_new_pages = end_pages - start_pages
+
+                months = max(period / 30.0, 1.0)
+                velocity.new_pages_per_month = round(
+                    velocity.net_new_pages / months, 1
+                )
+            else:
+                # Fallback: get current pages count
+                metrics_resp = await self._call_ahrefs(
+                    "site-explorer-metrics", {"target": domain}
+                )
+                velocity.total_pages_end = int(
+                    metrics_resp.get("data", {}).get("pages", 0)
+                )
+
+            velocities.append(velocity)
+            self.logger.info(
+                f"Content velocity for {domain}: "
+                f"{velocity.new_pages_per_month:.1f} pages/month"
+            )
+
+        return velocities
+
+    def generate_alerts(
+        self,
+        traffic_trends: list[TrafficTrend],
+        dr_trends: list[DrTrend],
+        keyword_movements: list[KeywordMovement],
+        content_velocities: list[ContentVelocity],
+        target_domain: str,
+    ) -> list[CompetitiveAlert]:
+        """Generate alerts for significant competitive movements."""
+        alerts: list[CompetitiveAlert] = []
+        target_domain = self._extract_domain(target_domain)
+
+        for trend in traffic_trends:
+            if trend.domain == target_domain:
+                continue
+            abs_growth = abs(trend.growth_rate) / 100.0
+            if abs_growth >= self.TRAFFIC_CHANGE_THRESHOLD:
+                severity = "critical" if abs_growth >= 0.50 else "warning"
+                direction = "surge" if trend.growth_rate > 0 else "decline"
+                alerts.append(CompetitiveAlert(
+                    domain=trend.domain,
+                    alert_type=f"traffic_{direction}",
+                    message=(
+                        f"{trend.domain} traffic {direction}: "
+                        f"{trend.growth_rate:+.1f}% "
+                        f"({trend.period_start_traffic:,} -> {trend.current_traffic:,})"
+                    ),
+                    severity=severity,
+                    metric_value=trend.growth_rate,
+                    threshold=self.TRAFFIC_CHANGE_THRESHOLD * 100,
+                ))
+
+        for trend in dr_trends:
+            if trend.domain == target_domain:
+                continue
+            if abs(trend.change) >= self.DR_CHANGE_THRESHOLD:
+                severity = "warning" if abs(trend.change) < 5 else "critical"
+                direction = "jump" if trend.change > 0 else "drop"
+                alerts.append(CompetitiveAlert(
+                    domain=trend.domain,
+                    alert_type=f"dr_{direction}",
+                    message=(
+                        f"{trend.domain} DR {direction}: "
+                        f"{trend.change:+.1f} points "
+                        f"({trend.period_start_dr:.1f} -> {trend.current_dr:.1f})"
+                    ),
+                    severity=severity,
+                    metric_value=trend.change,
+                    threshold=self.DR_CHANGE_THRESHOLD,
+                ))
+
+        for movement in keyword_movements:
+            if movement.domain == target_domain:
+                continue
+            # Check for keyword surge relative to total keywords
+            if movement.new_keywords > 0:
+                total_est = max(movement.new_keywords + movement.lost_keywords, 1)
+                surge_ratio = movement.new_keywords / total_est
+                if surge_ratio >= self.KEYWORD_SURGE_THRESHOLD:
+                    alerts.append(CompetitiveAlert(
+                        domain=movement.domain,
+                        alert_type="keyword_surge",
+                        message=(
+                            f"{movement.domain} gained {movement.new_keywords} "
+                            f"new keywords (net: {movement.net_change:+d})"
+                        ),
+                        severity="warning",
+                        metric_value=float(movement.new_keywords),
+                        threshold=self.KEYWORD_SURGE_THRESHOLD * 100,
+                    ))
+
+        # Check for content burst
+        if content_velocities:
+            avg_velocity = (
+                sum(v.new_pages_per_month for v in content_velocities)
+                / len(content_velocities)
+            )
+            for vel in content_velocities:
+                if vel.domain == target_domain:
+                    continue
+                if avg_velocity > 0 and vel.new_pages_per_month > avg_velocity * self.CONTENT_BURST_THRESHOLD:
+                    alerts.append(CompetitiveAlert(
+                        domain=vel.domain,
+                        alert_type="content_burst",
+                        message=(
+                            f"{vel.domain} publishing {vel.new_pages_per_month:.0f} "
+                            f"pages/month ({self.CONTENT_BURST_THRESHOLD:.0f}x above average)"
+                        ),
+                        severity="info",
+                        metric_value=vel.new_pages_per_month,
+                        threshold=avg_velocity * self.CONTENT_BURST_THRESHOLD,
+                    ))
+
+        # Sort alerts by severity
+        severity_order = {"critical": 0, "warning": 1, "info": 2}
+        alerts.sort(key=lambda a: severity_order.get(a.severity, 3))
+
+        self.logger.info(f"Generated {len(alerts)} competitive alerts")
+        return alerts
+
+    def estimate_market_share(
+        self,
+        traffic_trends: list[TrafficTrend],
+        keyword_movements: list[KeywordMovement],
+    ) -> list[MarketShare]:
+        """Estimate market share based on organic traffic within competitive set."""
+        shares: list[MarketShare] = []
+
+        total_traffic = sum(t.current_traffic for t in traffic_trends) or 1
+        total_kw_est = sum(
+            max(m.new_keywords + abs(m.net_change), 1) for m in keyword_movements
+        ) or 1
+
+        kw_by_domain: dict[str, int] = {}
+        for m in keyword_movements:
+            kw_by_domain[m.domain] = max(m.new_keywords + abs(m.net_change), 1)
+
+        for trend in traffic_trends:
+            share = MarketShare(domain=trend.domain)
+            share.traffic_share_pct = round(
+                (trend.current_traffic / total_traffic) * 100, 2
+            )
+            kw_count = kw_by_domain.get(trend.domain, 1)
+            share.keyword_share_pct = round(
+                (kw_count / total_kw_est) * 100, 2
+            )
+            share.overall_share_pct = round(
+                share.traffic_share_pct * 0.7 + share.keyword_share_pct * 0.3, 2
+            )
+            shares.append(share)
+
+        shares.sort(key=lambda s: s.overall_share_pct, reverse=True)
+        return shares
+
+    async def monitor(
+        self,
+        target: str,
+        competitors: list[str] | None = None,
+        period: int = 30,
+        scope: str = "all",
+    ) -> CompetitiveMonitorResult:
+        """Orchestrate full competitive monitoring pipeline."""
+        timestamp = datetime.now().isoformat()
+        target_domain = self._extract_domain(target)
+        result = CompetitiveMonitorResult(
+            target=target_domain,
+            period_days=period,
+            scope=scope,
+            timestamp=timestamp,
+        )
+
+        try:
+            # Discover competitors if not provided
+            if competitors:
+                comp_domains = [self._extract_domain(c) for c in competitors]
+            else:
+                self.logger.info("Auto-discovering competitors...")
+                comp_domains = await self._discover_competitors(target, limit=10)
+
+            all_domains = [target_domain] + comp_domains
+            self.logger.info(
+                f"Monitoring {len(all_domains)} domains over {period} days"
+            )
+
+            # Traffic trends
+            if scope in ("all", "traffic"):
+                self.logger.info("Fetching traffic trends...")
+                result.traffic_trends = await self.get_traffic_trends(
+                    all_domains, period
+                )
+
+            # DR trends
+            if scope in ("all", "traffic"):
+                self.logger.info("Fetching DR trends...")
+                result.dr_trends = await self.get_dr_trends(all_domains, period)
+
+            # Keyword movements
+            if scope in ("all", "keywords"):
+                self.logger.info("Tracking keyword movements...")
+                result.keyword_movements = await self.track_keyword_movement(
+                    all_domains, period
+                )
+
+            # Content velocity
+            if scope in ("all", "content"):
+                self.logger.info("Comparing content velocity...")
+                result.content_velocities = await self.compare_content_velocity(
+                    all_domains, period
+                )
+
+            # Generate alerts
+            self.logger.info("Generating competitive alerts...")
+            result.alerts = self.generate_alerts(
+                traffic_trends=result.traffic_trends,
+                dr_trends=result.dr_trends,
+                keyword_movements=result.keyword_movements,
+                content_velocities=result.content_velocities,
+                target_domain=target_domain,
+            )
+
+            # Market share estimation
+            if result.traffic_trends and result.keyword_movements:
+                self.logger.info("Estimating market shares...")
+                result.market_shares = self.estimate_market_share(
+                    result.traffic_trends,
+                    result.keyword_movements,
+                )
+
+            self.logger.info(
+                f"Monitoring complete: {len(result.alerts)} alerts generated"
+            )
+
+        except Exception as e:
+            msg = f"Monitoring pipeline error: {e}"
+            self.logger.error(msg)
+            result.errors.append(msg)
+
+        return result
+
+
+# ---------------------------------------------------------------------------
+# Output helpers
+# ---------------------------------------------------------------------------
+
+def _format_text_report(result: CompetitiveMonitorResult) -> str:
+    """Format monitoring result as human-readable text report."""
+    lines: list[str] = []
+    lines.append("=" * 70)
+    lines.append("  COMPETITIVE MONITORING REPORT")
+    lines.append(f"  Target: {result.target}")
+    lines.append(f"  Period: {result.period_days} days | Scope: {result.scope}")
+    lines.append(f"  Generated: {result.timestamp}")
+    lines.append("=" * 70)
+
+    # Alerts section
+    if result.alerts:
+        lines.append("")
+        lines.append("--- ALERTS ---")
+        for alert in result.alerts:
+            icon = {"critical": "[!]", "warning": "[*]", "info": "[-]"}.get(
+                alert.severity, "[-]"
+            )
+            lines.append(f"  {icon} [{alert.severity.upper()}] {alert.message}")
+
+    # Traffic trends
+    if result.traffic_trends:
+        lines.append("")
+        lines.append("--- TRAFFIC TRENDS ---")
+        lines.append(f"  {'Domain':<30} {'Direction':>10} {'Growth':>10} {'Current':>12}")
+        lines.append("  " + "-" * 65)
+        for t in result.traffic_trends:
+            marker = "*" if t.domain == result.target else " "
+            lines.append(
+                f"  {marker}{t.domain:<29} {t.direction:>10} "
+                f"{t.growth_rate:>+9.1f}% {t.current_traffic:>11,}"
+            )
+
+    # DR trends
+    if result.dr_trends:
+        lines.append("")
+        lines.append("--- DOMAIN RATING TRENDS ---")
+        lines.append(f"  {'Domain':<30} {'Direction':>10} {'Change':>10} {'Current':>10}")
+        lines.append("  " + "-" * 63)
+        for d in result.dr_trends:
+            marker = "*" if d.domain == result.target else " "
+            lines.append(
+                f"  {marker}{d.domain:<29} {d.direction:>10} "
+                f"{d.change:>+9.1f} {d.current_dr:>9.1f}"
+            )
+
+    # Keyword movements
+    if result.keyword_movements:
+        lines.append("")
+        lines.append("--- KEYWORD MOVEMENTS ---")
+        lines.append(
+            f"  {'Domain':<30} {'New':>8} {'Lost':>8} {'Net':>10}"
+        )
+        lines.append("  " + "-" * 59)
+        for k in result.keyword_movements:
+            marker = "*" if k.domain == result.target else " "
+            lines.append(
+                f"  {marker}{k.domain:<29} {k.new_keywords:>8,} "
+                f"{k.lost_keywords:>8,} {k.net_change:>+9,}"
+            )
+
+    # Content velocity
+    if result.content_velocities:
+        lines.append("")
+        lines.append("--- CONTENT VELOCITY ---")
+        lines.append(
+            f"  {'Domain':<30} {'Pages/Mo':>10} {'Net New':>10} {'Total':>10}"
+        )
+        lines.append("  " + "-" * 63)
+        for v in result.content_velocities:
+            marker = "*" if v.domain == result.target else " "
+            lines.append(
+                f"  {marker}{v.domain:<29} {v.new_pages_per_month:>9.1f} "
+                f"{v.net_new_pages:>9,} {v.total_pages_end:>9,}"
+            )
+
+    # Market share
+    if result.market_shares:
+        lines.append("")
+        lines.append("--- ESTIMATED MARKET SHARE ---")
+        lines.append(
+            f"  {'Domain':<30} {'Traffic%':>10} {'Keywords%':>10} {'Overall%':>10}"
+        )
+        lines.append("  " + "-" * 63)
+        for s in result.market_shares:
+            marker = "*" if s.domain == result.target else " "
+            lines.append(
+                f"  {marker}{s.domain:<29} {s.traffic_share_pct:>9.1f}% "
+                f"{s.keyword_share_pct:>9.1f}% {s.overall_share_pct:>9.1f}%"
+            )
+
+    if result.errors:
+        lines.append("")
+        lines.append("--- ERRORS ---")
+        for err in result.errors:
+            lines.append(f"  - {err}")
+
+    lines.append("")
+    lines.append("  * = target domain")
+    lines.append("=" * 70)
+    return "\n".join(lines)
+
+
+def _serialize_result(result: CompetitiveMonitorResult) -> dict:
+    """Convert result to JSON-serializable dict."""
+    output = {
+        "target": result.target,
+        "period_days": result.period_days,
+        "scope": result.scope,
+        "traffic_trends": [asdict(t) for t in result.traffic_trends],
+        "dr_trends": [asdict(d) for d in result.dr_trends],
+        "keyword_movements": [asdict(k) for k in result.keyword_movements],
+        "content_velocities": [asdict(v) for v in result.content_velocities],
+        "alerts": [asdict(a) for a in result.alerts],
+        "market_shares": [asdict(s) for s in result.market_shares],
+        "timestamp": result.timestamp,
+    }
+    if result.errors:
+        output["errors"] = result.errors
+    return output
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def parse_args(argv: list[str] | None = None) -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Competitive Monitor - Track SEO competitive changes over time",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""\
+Examples:
+  python competitive_monitor.py --target https://example.com --period 30 --json
+  python competitive_monitor.py --target https://example.com --competitor https://comp1.com --period 60 --json
+  python competitive_monitor.py --target https://example.com --scope traffic --period 90 --json
+        """,
+    )
+    parser.add_argument(
+        "--target",
+        required=True,
+        help="Target website URL or domain to monitor",
+    )
+    parser.add_argument(
+        "--competitor",
+        action="append",
+        dest="competitors",
+        default=[],
+        help="Competitor URL/domain (repeatable; omit for auto-discovery)",
+    )
+    parser.add_argument(
+        "--period",
+        type=int,
+        default=30,
+        help="Monitoring period in days (default: 30)",
+    )
+    parser.add_argument(
+        "--scope",
+        choices=["all", "traffic", "keywords", "content"],
+        default="all",
+        help="Monitoring scope (default: all)",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        default=False,
+        help="Output in JSON format",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        default=None,
+        help="Save output to file path",
+    )
+    return parser.parse_args(argv)
+
+
+async def async_main(args: argparse.Namespace) -> None:
+    monitor = CompetitiveMonitor()
+
+    result = await monitor.monitor(
+        target=args.target,
+        competitors=args.competitors or None,
+        period=args.period,
+        scope=args.scope,
+    )
+
+    if args.json:
+        output_str = json.dumps(_serialize_result(result), indent=2, ensure_ascii=False)
+    else:
+        output_str = _format_text_report(result)
+
+    if args.output:
+        with open(args.output, "w", encoding="utf-8") as f:
+            f.write(output_str)
+        logger.info(f"Report saved to {args.output}")
+    else:
+        print(output_str)
+
+    monitor.print_stats()
+
+
+def main() -> None:
+    args = parse_args()
+    asyncio.run(async_main(args))
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/31-seo-competitor-intel/code/scripts/competitor_profiler.py b/custom-skills/31-seo-competitor-intel/code/scripts/competitor_profiler.py
new file mode 100644
index 0000000..0e6a883
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/code/scripts/competitor_profiler.py
@@ -0,0 +1,719 @@
+"""
+Competitor Profiler - SEO Competitive Intelligence
+===================================================
+Purpose: Auto-discover competitors, build profile cards, comparison matrices,
+         keyword overlap analysis, and competitive threat scoring.
+Python: 3.10+
+
+Usage:
+    python competitor_profiler.py --target https://example.com --json
+    python competitor_profiler.py --target https://example.com --competitor https://comp1.com --json
+    python competitor_profiler.py --target https://example.com --max-competitors 10 --korean-market --json
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import sys
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from typing import Any
+from urllib.parse import urlparse
+
+from base_client import BaseAsyncClient, config
+
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class CompetitorProfile:
+    """Full profile card for a single domain."""
+    domain: str
+    domain_rating: float = 0.0
+    organic_traffic: int = 0
+    organic_keywords: int = 0
+    referring_domains: int = 0
+    top_pages_count: int = 0
+    traffic_value_usd: float = 0.0
+    content_volume: int = 0
+    naver_blog_presence: bool = False
+    naver_cafe_presence: bool = False
+
+
+@dataclass
+class KeywordOverlap:
+    """Keyword overlap analysis between target and a competitor."""
+    shared: int = 0
+    unique_target: int = 0
+    unique_competitor: int = 0
+    gap_keywords: int = 0
+    overlap_percentage: float = 0.0
+
+
+@dataclass
+class ThreatAssessment:
+    """Competitive threat score and breakdown for one competitor."""
+    domain: str = ""
+    threat_score: float = 0.0
+    growth_rate: float = 0.0
+    dr_gap: float = 0.0
+    keyword_overlap_pct: float = 0.0
+    traffic_ratio: float = 0.0
+    strengths: list[str] = field(default_factory=list)
+    weaknesses: list[str] = field(default_factory=list)
+
+
+@dataclass
+class ComparisonMatrix:
+    """Multi-dimensional comparison matrix across SEO dimensions."""
+    dimensions: list[str] = field(default_factory=list)
+    target_scores: dict[str, float] = field(default_factory=dict)
+    competitor_scores: dict[str, dict[str, float]] = field(default_factory=dict)
+
+
+@dataclass
+class CompetitorProfilingResult:
+    """Full profiling result with all competitor data."""
+    target: str = ""
+    target_profile: CompetitorProfile | None = None
+    competitors: list[dict[str, Any]] = field(default_factory=list)
+    comparison_matrix: ComparisonMatrix | None = None
+    market_position: str = "unknown"
+    timestamp: str = ""
+    errors: list[str] = field(default_factory=list)
+
+
+# ---------------------------------------------------------------------------
+# Profiler
+# ---------------------------------------------------------------------------
+
+class CompetitorProfiler(BaseAsyncClient):
+    """Builds competitor profiles using Ahrefs MCP tools."""
+
+    DIMENSIONS = ["traffic", "domain_rating", "keywords", "backlinks", "content"]
+
+    def __init__(self, korean_market: bool = False):
+        super().__init__(max_concurrent=5, requests_per_second=2.0)
+        self.korean_market = korean_market
+
+    @staticmethod
+    def _extract_domain(url: str) -> str:
+        """Extract bare domain from URL or return as-is if already bare."""
+        if "://" in url:
+            parsed = urlparse(url)
+            return parsed.netloc.lower().replace("www.", "")
+        return url.lower().replace("www.", "")
+
+    # ------------------------------------------------------------------
+    # Ahrefs MCP wrappers (return dicts; Claude MCP bridge fills these)
+    # ------------------------------------------------------------------
+
+    async def _call_ahrefs(self, tool: str, params: dict[str, Any]) -> dict:
+        """Simulate Ahrefs MCP call. In production, routed via MCP bridge."""
+        self.logger.info(f"Ahrefs MCP call: {tool} | params={params}")
+        return {"tool": tool, "params": params, "data": {}}
+
+    async def discover_competitors(
+        self, target: str, limit: int = 20
+    ) -> list[str]:
+        """Discover organic competitors via site-explorer-organic-competitors."""
+        domain = self._extract_domain(target)
+        self.logger.info(f"Discovering competitors for {domain} (limit={limit})")
+
+        resp = await self._call_ahrefs(
+            "site-explorer-organic-competitors",
+            {"target": domain, "limit": limit, "country": "kr"},
+        )
+
+        competitors_raw: list[dict] = resp.get("data", {}).get("competitors", [])
+        discovered = []
+        for entry in competitors_raw:
+            comp_domain = entry.get("domain", "")
+            if comp_domain and comp_domain != domain:
+                discovered.append(comp_domain)
+
+        if not discovered:
+            self.logger.warning(
+                "No competitors returned from Ahrefs; "
+                "check that the target domain has organic traffic."
+            )
+        else:
+            self.logger.info(f"Discovered {len(discovered)} competitors")
+
+        return discovered[:limit]
+
+    async def build_profile(self, domain: str) -> CompetitorProfile:
+        """Build a complete profile card for a single domain."""
+        domain = self._extract_domain(domain)
+        profile = CompetitorProfile(domain=domain)
+
+        # --- Metrics ---
+        metrics_resp = await self._call_ahrefs(
+            "site-explorer-metrics", {"target": domain}
+        )
+        metrics = metrics_resp.get("data", {})
+        profile.organic_traffic = int(metrics.get("organic_traffic", 0))
+        profile.organic_keywords = int(metrics.get("organic_keywords", 0))
+        profile.traffic_value_usd = float(metrics.get("traffic_value", 0.0))
+
+        # --- Domain Rating ---
+        dr_resp = await self._call_ahrefs(
+            "site-explorer-domain-rating", {"target": domain}
+        )
+        dr_data = dr_resp.get("data", {})
+        profile.domain_rating = float(dr_data.get("domain_rating", 0.0))
+
+        # --- Referring Domains ---
+        bl_resp = await self._call_ahrefs(
+            "site-explorer-backlinks-stats", {"target": domain}
+        )
+        bl_data = bl_resp.get("data", {})
+        profile.referring_domains = int(bl_data.get("referring_domains", 0))
+
+        # --- Top Pages ---
+        pages_resp = await self._call_ahrefs(
+            "site-explorer-top-pages", {"target": domain, "limit": 1000}
+        )
+        pages_data = pages_resp.get("data", {})
+        profile.top_pages_count = len(pages_data.get("pages", []))
+
+        # --- Content Volume (pages indexed) ---
+        history_resp = await self._call_ahrefs(
+            "site-explorer-pages-history", {"target": domain}
+        )
+        history_data = history_resp.get("data", {})
+        data_points = history_data.get("data_points", [])
+        if data_points:
+            latest = data_points[-1]
+            profile.content_volume = int(latest.get("pages", 0))
+
+        self.logger.info(
+            f"Profile built for {domain}: DR={profile.domain_rating}, "
+            f"traffic={profile.organic_traffic}, keywords={profile.organic_keywords}"
+        )
+        return profile
+
+    async def analyze_keyword_overlap(
+        self, target: str, competitor: str, limit: int = 1000
+    ) -> KeywordOverlap:
+        """Analyze keyword overlap between target and a single competitor."""
+        target_domain = self._extract_domain(target)
+        comp_domain = self._extract_domain(competitor)
+
+        # Fetch keyword sets for both domains
+        target_resp = await self._call_ahrefs(
+            "site-explorer-organic-keywords",
+            {"target": target_domain, "limit": limit},
+        )
+        comp_resp = await self._call_ahrefs(
+            "site-explorer-organic-keywords",
+            {"target": comp_domain, "limit": limit},
+        )
+
+        target_kws: set[str] = set()
+        for kw in target_resp.get("data", {}).get("keywords", []):
+            keyword = kw.get("keyword", "")
+            if keyword:
+                target_kws.add(keyword.lower())
+
+        comp_kws: set[str] = set()
+        for kw in comp_resp.get("data", {}).get("keywords", []):
+            keyword = kw.get("keyword", "")
+            if keyword:
+                comp_kws.add(keyword.lower())
+
+        shared = target_kws & comp_kws
+        unique_target = target_kws - comp_kws
+        unique_comp = comp_kws - target_kws
+        gap = unique_comp  # keywords the competitor ranks for but target does not
+
+        total_union = len(target_kws | comp_kws) or 1
+        overlap_pct = (len(shared) / total_union) * 100.0
+
+        overlap = KeywordOverlap(
+            shared=len(shared),
+            unique_target=len(unique_target),
+            unique_competitor=len(unique_comp),
+            gap_keywords=len(gap),
+            overlap_percentage=round(overlap_pct, 2),
+        )
+        self.logger.info(
+            f"Keyword overlap {target_domain} vs {comp_domain}: "
+            f"shared={overlap.shared}, gap={overlap.gap_keywords}"
+        )
+        return overlap
+
+    def build_comparison_matrix(
+        self,
+        target_profile: CompetitorProfile,
+        competitor_profiles: list[CompetitorProfile],
+    ) -> ComparisonMatrix:
+        """Create a multi-dimensional comparison matrix."""
+        matrix = ComparisonMatrix(dimensions=list(self.DIMENSIONS))
+
+        # Normalize scores to 0-100 scale relative to max in competitive set
+        all_profiles = [target_profile] + competitor_profiles
+
+        def _max_val(attr: str) -> float:
+            return max(getattr(p, attr, 0) for p in all_profiles) or 1
+
+        max_traffic = _max_val("organic_traffic")
+        max_dr = 100.0  # DR is already 0-100
+        max_kw = _max_val("organic_keywords")
+        max_rd = _max_val("referring_domains")
+        max_content = _max_val("content_volume")
+
+        def _norm(profile: CompetitorProfile) -> dict[str, float]:
+            return {
+                "traffic": round((profile.organic_traffic / max_traffic) * 100, 1),
+                "domain_rating": round(profile.domain_rating, 1),
+                "keywords": round((profile.organic_keywords / max_kw) * 100, 1),
+                "backlinks": round((profile.referring_domains / max_rd) * 100, 1),
+                "content": round((profile.content_volume / max_content) * 100, 1)
+                if max_content > 0
+                else 0.0,
+            }
+
+        matrix.target_scores = _norm(target_profile)
+
+        for cp in competitor_profiles:
+            matrix.competitor_scores[cp.domain] = _norm(cp)
+
+        return matrix
+
+    def score_threat(
+        self,
+        target_profile: CompetitorProfile,
+        competitor_profile: CompetitorProfile,
+        overlap: KeywordOverlap,
+    ) -> ThreatAssessment:
+        """Score competitive threat 0-100 based on multiple factors."""
+        assessment = ThreatAssessment(domain=competitor_profile.domain)
+
+        # --- DR gap (positive = competitor stronger) ---
+        dr_gap = competitor_profile.domain_rating - target_profile.domain_rating
+        assessment.dr_gap = round(dr_gap, 1)
+        dr_score = min(max((dr_gap + 30) / 60 * 100, 0), 100)  # scale -30..+30 -> 0-100
+
+        # --- Traffic ratio ---
+        target_traffic = max(target_profile.organic_traffic, 1)
+        traffic_ratio = competitor_profile.organic_traffic / target_traffic
+        assessment.traffic_ratio = round(traffic_ratio, 2)
+        traffic_score = min(traffic_ratio * 50, 100)  # 2x traffic = 100
+
+        # --- Keyword overlap percentage ---
+        assessment.keyword_overlap_pct = overlap.overlap_percentage
+        overlap_score = min(overlap.overlap_percentage * 2, 100)  # 50% overlap = 100
+
+        # --- Gap keywords (competitor ranks, target doesn't) ---
+        total_target_kw = max(overlap.shared + overlap.unique_target, 1)
+        gap_ratio = overlap.gap_keywords / total_target_kw
+        gap_score = min(gap_ratio * 100, 100)
+
+        # --- Weighted threat score ---
+        threat = (
+            dr_score * 0.20
+            + traffic_score * 0.30
+            + overlap_score * 0.25
+            + gap_score * 0.25
+        )
+        assessment.threat_score = round(min(max(threat, 0), 100), 1)
+
+        # --- Identify strengths & weaknesses ---
+        if dr_gap > 5:
+            assessment.strengths.append(f"Higher DR by {dr_gap:.0f} points")
+        elif dr_gap < -5:
+            assessment.weaknesses.append(f"Lower DR by {abs(dr_gap):.0f} points")
+
+        if traffic_ratio > 1.5:
+            assessment.strengths.append(
+                f"Traffic {traffic_ratio:.1f}x higher than target"
+            )
+        elif traffic_ratio < 0.5:
+            assessment.weaknesses.append(
+                f"Traffic only {traffic_ratio:.1f}x of target"
+            )
+
+        if overlap.gap_keywords > overlap.shared:
+            assessment.strengths.append(
+                f"{overlap.gap_keywords} keywords target is missing"
+            )
+
+        if competitor_profile.referring_domains > target_profile.referring_domains * 1.5:
+            assessment.strengths.append("Significantly more referring domains")
+        elif competitor_profile.referring_domains < target_profile.referring_domains * 0.5:
+            assessment.weaknesses.append("Fewer referring domains")
+
+        if competitor_profile.content_volume > target_profile.content_volume * 1.5:
+            assessment.strengths.append("Larger content volume")
+        elif competitor_profile.content_volume < target_profile.content_volume * 0.5:
+            assessment.weaknesses.append("Smaller content library")
+
+        self.logger.info(
+            f"Threat score for {competitor_profile.domain}: "
+            f"{assessment.threat_score}/100"
+        )
+        return assessment
+
+    async def detect_korean_presence(self, domain: str) -> dict[str, bool]:
+        """Check Naver Blog/Cafe presence for a domain (heuristic)."""
+        domain = self._extract_domain(domain)
+        self.logger.info(f"Checking Korean market presence for {domain}")
+
+        # In production, this would use WebSearch MCP to query Naver
+        # Heuristic: check if domain has .co.kr or .kr TLD,
+        # or has Korean-language top pages
+        is_korean_tld = domain.endswith(".kr") or domain.endswith(".co.kr")
+
+        # Check top pages for Korean content signals
+        pages_resp = await self._call_ahrefs(
+            "site-explorer-organic-keywords",
+            {"target": domain, "limit": 50, "country": "kr"},
+        )
+        kr_keywords = pages_resp.get("data", {}).get("keywords", [])
+        has_kr_keywords = len(kr_keywords) > 0
+
+        return {
+            "naver_blog_presence": is_korean_tld or has_kr_keywords,
+            "naver_cafe_presence": is_korean_tld,
+            "korean_tld": is_korean_tld,
+            "korean_keyword_count": len(kr_keywords),
+        }
+
+    def determine_market_position(
+        self,
+        target_profile: CompetitorProfile,
+        competitor_profiles: list[CompetitorProfile],
+    ) -> str:
+        """Classify target as leader / challenger / follower / niche."""
+        if not competitor_profiles:
+            return "unknown"
+
+        all_profiles = [target_profile] + competitor_profiles
+        all_profiles.sort(key=lambda p: p.organic_traffic, reverse=True)
+
+        target_rank = next(
+            (i for i, p in enumerate(all_profiles) if p.domain == target_profile.domain),
+            len(all_profiles),
+        )
+
+        total = len(all_profiles)
+        percentile = target_rank / total
+
+        # DR comparison
+        avg_competitor_dr = (
+            sum(p.domain_rating for p in competitor_profiles) / len(competitor_profiles)
+            if competitor_profiles
+            else 0
+        )
+        dr_advantage = target_profile.domain_rating - avg_competitor_dr
+
+        # Traffic leader check
+        max_traffic = max(p.organic_traffic for p in all_profiles) or 1
+        traffic_share = target_profile.organic_traffic / max_traffic
+
+        if percentile <= 0.1 and traffic_share >= 0.8:
+            return "leader"
+        elif percentile <= 0.33 or (dr_advantage > 10 and traffic_share > 0.5):
+            return "challenger"
+        elif percentile <= 0.66:
+            return "follower"
+        else:
+            # Check if niche player (high DR but low traffic = niche authority)
+            if target_profile.domain_rating > avg_competitor_dr:
+                return "niche"
+            return "follower"
+
+    async def profile(
+        self,
+        target: str,
+        competitors: list[str] | None = None,
+        max_competitors: int = 10,
+    ) -> CompetitorProfilingResult:
+        """Orchestrate full competitor profiling pipeline."""
+        timestamp = datetime.now().isoformat()
+        result = CompetitorProfilingResult(
+            target=self._extract_domain(target),
+            timestamp=timestamp,
+        )
+
+        try:
+            # Step 1: Build target profile
+            self.logger.info("Step 1/6: Building target profile...")
+            target_profile = await self.build_profile(target)
+            result.target_profile = target_profile
+
+            # Step 2: Discover or validate competitors
+            self.logger.info("Step 2/6: Discovering competitors...")
+            if competitors:
+                comp_domains = [self._extract_domain(c) for c in competitors]
+            else:
+                comp_domains = await self.discover_competitors(
+                    target, limit=max_competitors
+                )
+
+            if not comp_domains:
+                result.errors.append("No competitors found or provided.")
+                return result
+
+            comp_domains = comp_domains[:max_competitors]
+
+            # Step 3: Build competitor profiles
+            self.logger.info(
+                f"Step 3/6: Profiling {len(comp_domains)} competitors..."
+            )
+            competitor_profiles: list[CompetitorProfile] = []
+            for domain in comp_domains:
+                try:
+                    cp = await self.build_profile(domain)
+                    if self.korean_market:
+                        kr_presence = await self.detect_korean_presence(domain)
+                        cp.naver_blog_presence = kr_presence.get(
+                            "naver_blog_presence", False
+                        )
+                        cp.naver_cafe_presence = kr_presence.get(
+                            "naver_cafe_presence", False
+                        )
+                    competitor_profiles.append(cp)
+                except Exception as e:
+                    msg = f"Failed to profile {domain}: {e}"
+                    self.logger.error(msg)
+                    result.errors.append(msg)
+
+            # Step 4: Keyword overlap analysis
+            self.logger.info("Step 4/6: Analyzing keyword overlaps...")
+            overlaps: dict[str, KeywordOverlap] = {}
+            for cp in competitor_profiles:
+                try:
+                    overlap = await self.analyze_keyword_overlap(target, cp.domain)
+                    overlaps[cp.domain] = overlap
+                except Exception as e:
+                    msg = f"Keyword overlap failed for {cp.domain}: {e}"
+                    self.logger.error(msg)
+                    result.errors.append(msg)
+                    overlaps[cp.domain] = KeywordOverlap()
+
+            # Step 5: Build comparison matrix
+            self.logger.info("Step 5/6: Building comparison matrix...")
+            matrix = self.build_comparison_matrix(target_profile, competitor_profiles)
+            result.comparison_matrix = matrix
+
+            # Step 6: Score threats and assemble output
+            self.logger.info("Step 6/6: Scoring competitive threats...")
+            for cp in competitor_profiles:
+                overlap = overlaps.get(cp.domain, KeywordOverlap())
+                threat = self.score_threat(target_profile, cp, overlap)
+
+                competitor_entry = {
+                    "domain": cp.domain,
+                    "profile": asdict(cp),
+                    "threat_score": threat.threat_score,
+                    "threat_detail": asdict(threat),
+                    "keyword_overlap": asdict(overlap),
+                }
+                result.competitors.append(competitor_entry)
+
+            # Sort by threat score descending
+            result.competitors.sort(
+                key=lambda c: c.get("threat_score", 0), reverse=True
+            )
+
+            # Determine market position
+            result.market_position = self.determine_market_position(
+                target_profile, competitor_profiles
+            )
+
+            self.logger.info(
+                f"Profiling complete: {len(result.competitors)} competitors analyzed. "
+                f"Market position: {result.market_position}"
+            )
+
+        except Exception as e:
+            msg = f"Profiling pipeline error: {e}"
+            self.logger.error(msg)
+            result.errors.append(msg)
+
+        return result
+
+
+# ---------------------------------------------------------------------------
+# Output helpers
+# ---------------------------------------------------------------------------
+
+def _format_text_report(result: CompetitorProfilingResult) -> str:
+    """Format profiling result as human-readable text report."""
+    lines: list[str] = []
+    lines.append("=" * 70)
+    lines.append(f"  COMPETITOR INTELLIGENCE REPORT")
+    lines.append(f"  Target: {result.target}")
+    lines.append(f"  Generated: {result.timestamp}")
+    lines.append(f"  Market Position: {result.market_position.upper()}")
+    lines.append("=" * 70)
+
+    if result.target_profile:
+        tp = result.target_profile
+        lines.append("")
+        lines.append("--- TARGET PROFILE ---")
+        lines.append(f"  Domain Rating:     {tp.domain_rating}")
+        lines.append(f"  Organic Traffic:   {tp.organic_traffic:,}")
+        lines.append(f"  Organic Keywords:  {tp.organic_keywords:,}")
+        lines.append(f"  Referring Domains: {tp.referring_domains:,}")
+        lines.append(f"  Top Pages:         {tp.top_pages_count:,}")
+        lines.append(f"  Content Volume:    {tp.content_volume:,}")
+        lines.append(f"  Traffic Value:     ${tp.traffic_value_usd:,.2f}")
+
+    if result.competitors:
+        lines.append("")
+        lines.append("--- COMPETITORS (sorted by threat score) ---")
+        for i, comp in enumerate(result.competitors, 1):
+            p = comp["profile"]
+            t = comp["threat_detail"]
+            o = comp["keyword_overlap"]
+            lines.append("")
+            lines.append(f"  #{i} {comp['domain']}")
+            lines.append(f"     Threat Score:    {comp['threat_score']}/100")
+            lines.append(f"     Domain Rating:   {p['domain_rating']}")
+            lines.append(f"     Organic Traffic: {p['organic_traffic']:,}")
+            lines.append(f"     Keywords:        {p['organic_keywords']:,}")
+            lines.append(f"     Referring Doms:  {p['referring_domains']:,}")
+            lines.append(f"     Keyword Overlap: {o['shared']} shared, {o['gap_keywords']} gap")
+            if t.get("strengths"):
+                lines.append(f"     Strengths:       {'; '.join(t['strengths'])}")
+            if t.get("weaknesses"):
+                lines.append(f"     Weaknesses:      {'; '.join(t['weaknesses'])}")
+
+    if result.comparison_matrix:
+        m = result.comparison_matrix
+        lines.append("")
+        lines.append("--- COMPARISON MATRIX ---")
+        header = f"  {'Dimension':<20} {'Target':>10}"
+        for domain in m.competitor_scores:
+            short = domain[:15]
+            header += f" {short:>15}"
+        lines.append(header)
+        lines.append("  " + "-" * (len(header) - 2))
+
+        for dim in m.dimensions:
+            row = f"  {dim:<20} {m.target_scores.get(dim, 0):>10.1f}"
+            for domain, scores in m.competitor_scores.items():
+                row += f" {scores.get(dim, 0):>15.1f}"
+            lines.append(row)
+
+    if result.errors:
+        lines.append("")
+        lines.append("--- ERRORS ---")
+        for err in result.errors:
+            lines.append(f"  - {err}")
+
+    lines.append("")
+    lines.append("=" * 70)
+    return "\n".join(lines)
+
+
+def _serialize_result(result: CompetitorProfilingResult) -> dict:
+    """Convert result to JSON-serializable dict."""
+    output = {
+        "target": result.target,
+        "target_profile": asdict(result.target_profile) if result.target_profile else None,
+        "competitors": result.competitors,
+        "comparison_matrix": asdict(result.comparison_matrix)
+        if result.comparison_matrix
+        else None,
+        "market_position": result.market_position,
+        "timestamp": result.timestamp,
+    }
+    if result.errors:
+        output["errors"] = result.errors
+    return output
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def parse_args(argv: list[str] | None = None) -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="SEO Competitor Profiler - Build competitive intelligence reports",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""\
+Examples:
+  python competitor_profiler.py --target https://example.com --json
+  python competitor_profiler.py --target https://example.com --competitor https://comp1.com --json
+  python competitor_profiler.py --target https://example.com --max-competitors 10 --korean-market --json
+        """,
+    )
+    parser.add_argument(
+        "--target",
+        required=True,
+        help="Target website URL or domain to analyze",
+    )
+    parser.add_argument(
+        "--competitor",
+        action="append",
+        dest="competitors",
+        default=[],
+        help="Competitor URL/domain (repeatable; omit for auto-discovery)",
+    )
+    parser.add_argument(
+        "--max-competitors",
+        type=int,
+        default=10,
+        help="Maximum competitors to profile (default: 10)",
+    )
+    parser.add_argument(
+        "--korean-market",
+        action="store_true",
+        default=False,
+        help="Include Korean market analysis (Naver Blog/Cafe presence)",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        default=False,
+        help="Output in JSON format",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        default=None,
+        help="Save output to file path",
+    )
+    return parser.parse_args(argv)
+
+
+async def async_main(args: argparse.Namespace) -> None:
+    profiler = CompetitorProfiler(korean_market=args.korean_market)
+
+    result = await profiler.profile(
+        target=args.target,
+        competitors=args.competitors or None,
+        max_competitors=args.max_competitors,
+    )
+
+    if args.json:
+        output_str = json.dumps(_serialize_result(result), indent=2, ensure_ascii=False)
+    else:
+        output_str = _format_text_report(result)
+
+    if args.output:
+        with open(args.output, "w", encoding="utf-8") as f:
+            f.write(output_str)
+        logger.info(f"Report saved to {args.output}")
+    else:
+        print(output_str)
+
+    profiler.print_stats()
+
+
+def main() -> None:
+    args = parse_args()
+    asyncio.run(async_main(args))
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/31-seo-competitor-intel/code/scripts/requirements.txt b/custom-skills/31-seo-competitor-intel/code/scripts/requirements.txt
new file mode 100644
index 0000000..1063c35
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/code/scripts/requirements.txt
@@ -0,0 +1,8 @@
+# 31-seo-competitor-intel dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+pandas>=2.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/31-seo-competitor-intel/desktop/SKILL.md b/custom-skills/31-seo-competitor-intel/desktop/SKILL.md
new file mode 100644
index 0000000..be468a3
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/desktop/SKILL.md
@@ -0,0 +1,147 @@
+---
+name: seo-competitor-intel
+description: |
+  Competitor intelligence and SEO benchmarking. Triggers: competitor analysis, competitive intelligence, competitor comparison, threat assessment, market position, benchmarking.
+---
+
+# SEO Competitor Intelligence & Benchmarking
+
+## Purpose
+
+Comprehensive competitor intelligence for SEO: auto-discover competitors, build profile cards, create head-to-head comparison matrices, analyze keyword overlap, track traffic trends, and score competitive threats. Supports Korean market analysis including Naver Blog/Cafe presence.
+
+## Core Capabilities
+
+1. **Competitor Discovery** - Auto-discover organic competitors via Ahrefs
+2. **Profile Cards** - DR, traffic, keywords, referring domains, top pages, content volume
+3. **Comparison Matrix** - Multi-dimensional head-to-head comparison
+4. **Keyword Overlap** - Shared, unique, and gap keyword analysis
+5. **Threat Scoring** - 0-100 score based on DR gap, traffic ratio, keyword overlap, growth
+6. **Competitive Monitoring** - Traffic trends, DR changes, keyword movement, content velocity
+7. **Alert Generation** - Flag significant competitive movements
+8. **Market Share Estimation** - Organic traffic share within competitive set
+
+## MCP Tool Usage
+
+### Ahrefs for Competitive Data
+```
+mcp__ahrefs__site-explorer-organic-competitors: Discover organic competitors
+mcp__ahrefs__site-explorer-metrics: Current organic metrics
+mcp__ahrefs__site-explorer-metrics-history: Historical metric trends
+mcp__ahrefs__site-explorer-domain-rating: Domain Rating score
+mcp__ahrefs__site-explorer-domain-rating-history: DR trend over time
+mcp__ahrefs__site-explorer-organic-keywords: Keyword rankings per domain
+mcp__ahrefs__site-explorer-top-pages: Top performing pages
+mcp__ahrefs__site-explorer-pages-by-traffic: Pages ranked by traffic
+mcp__ahrefs__site-explorer-referring-domains: Referring domain list
+mcp__ahrefs__site-explorer-backlinks-stats: Backlink overview
+mcp__ahrefs__site-explorer-pages-history: Page index history
+```
+
+### Notion for Report Storage
+```
+mcp__notion__notion-create-pages: Save reports to SEO Audit Log
+```
+
+### WebSearch for Korean Market
+```
+WebSearch: Check Naver Blog/Cafe presence for competitors
+```
+
+## Workflow
+
+### Competitor Profiling
+1. Accept target URL/domain
+2. Auto-discover competitors via Ahrefs organic-competitors (or use provided list)
+3. Build profile card for target and each competitor (DR, traffic, keywords, backlinks, content)
+4. Analyze keyword overlap between target and each competitor
+5. Build multi-dimensional comparison matrix
+6. Score competitive threats (0-100)
+7. Determine market position (leader/challenger/follower/niche)
+8. If Korean market: check Naver Blog/Cafe presence
+
+### Competitive Monitoring
+1. Accept target, competitors, and monitoring period
+2. Fetch traffic trend history for all domains
+3. Fetch DR trend history for all domains
+4. Track keyword movement (new/lost keywords)
+5. Compare content publication velocity
+6. Generate alerts for significant changes (>20% traffic, DR jump, keyword surge)
+7. Estimate market share within competitive set
+
+## Output Format
+
+### Profiling Report
+```markdown
+## Competitor Intelligence Report: [domain]
+
+### Target Profile
+- Domain Rating: [DR]
+- Organic Traffic: [traffic]
+- Keywords: [count]
+- Referring Domains: [count]
+
+### Competitors (by threat score)
+1. **[competitor.com]** - Threat: [score]/100
+   - DR: [value] | Traffic: [value] | Keywords: [value]
+   - Keyword Overlap: [shared] shared, [gap] gap
+   - Strengths: [list]
+   - Weaknesses: [list]
+
+### Comparison Matrix
+| Dimension | Target | Comp1 | Comp2 |
+|-----------|--------|-------|-------|
+
+### Market Position: [leader/challenger/follower/niche]
+```
+
+### Monitoring Report
+```markdown
+## Competitive Monitoring Report: [domain]
+### Period: [N] days
+
+### Alerts
+- [severity] [message]
+
+### Traffic Trends
+| Domain | Direction | Growth | Current |
+
+### Keyword Movements
+| Domain | New | Lost | Net |
+
+### Market Share
+| Domain | Traffic% | Overall% |
+```
+
+## Threat Scoring Methodology
+
+| Factor | Weight | Scale |
+|--------|--------|-------|
+| DR Gap | 20% | -30 to +30 mapped to 0-100 |
+| Traffic Ratio | 30% | 0x to 2x+ mapped to 0-100 |
+| Keyword Overlap | 25% | 0-50%+ mapped to 0-100 |
+| Gap Keywords | 25% | Ratio to target keywords |
+
+## Alert Thresholds
+
+| Alert Type | Threshold | Severity |
+|------------|-----------|----------|
+| Traffic change | >20% | warning; >50% critical |
+| DR change | >3 points | warning; >5 critical |
+| Keyword surge | >15% growth | warning |
+| Content burst | >2x avg velocity | info |
+
+## Limitations
+
+- Ahrefs data has ~24h freshness lag
+- Keyword overlap limited to top 1,000 keywords per domain
+- Content velocity based on Ahrefs page index (not real-time crawl)
+- Naver presence detection is heuristic-based
+
+## Notion Output (Required)
+
+All reports MUST be saved to OurDigital SEO Audit Log:
+- **Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **Properties**: Issue (title), Site (url), Category ("Competitor Intelligence"), Priority, Found Date, Audit ID
+- **Language**: Korean with English technical terms
+- **Audit ID Format**: COMP-YYYYMMDD-NNN
diff --git a/custom-skills/31-seo-competitor-intel/desktop/skill.yaml b/custom-skills/31-seo-competitor-intel/desktop/skill.yaml
new file mode 100644
index 0000000..c55c3ff
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-competitor-intel
+description: |
+  Competitor intelligence and SEO benchmarking. Triggers: competitor analysis, competitive intelligence, competitor comparison, threat assessment, market position, benchmarking.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/31-seo-competitor-intel/desktop/tools/ahrefs.md b/custom-skills/31-seo-competitor-intel/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..ba979df
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/desktop/tools/ahrefs.md
@@ -0,0 +1,35 @@
+# Ahrefs
+
+> MCP tool documentation for competitor intelligence skill
+
+## Available Commands
+
+- `site-explorer-metrics` - Get current organic metrics for a domain
+- `site-explorer-metrics-history` - Get historical organic metrics trends
+- `site-explorer-organic-competitors` - Discover organic search competitors
+- `site-explorer-organic-keywords` - Get keyword rankings per domain
+- `site-explorer-top-pages` - Get top performing pages by traffic
+- `site-explorer-pages-by-traffic` - Get pages ranked by organic traffic
+- `site-explorer-domain-rating` - Get Domain Rating score
+- `site-explorer-domain-rating-history` - Get DR trend over time
+- `site-explorer-referring-domains` - Get list of referring domains
+- `site-explorer-backlinks-stats` - Get backlink overview statistics
+- `site-explorer-pages-history` - Get page index history
+
+## Configuration
+
+- Requires Ahrefs MCP server configured in Claude Desktop
+- API access via `mcp__ahrefs__*` tool prefix
+
+## Examples
+
+```
+# Discover competitors
+mcp__ahrefs__site-explorer-organic-competitors(target="example.com")
+
+# Get domain metrics
+mcp__ahrefs__site-explorer-metrics(target="example.com")
+
+# Get keyword rankings
+mcp__ahrefs__site-explorer-organic-keywords(target="example.com", limit=100)
+```
diff --git a/custom-skills/31-seo-competitor-intel/desktop/tools/notion.md b/custom-skills/31-seo-competitor-intel/desktop/tools/notion.md
new file mode 100644
index 0000000..8ef7c81
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/desktop/tools/notion.md
@@ -0,0 +1,33 @@
+# Notion
+
+> MCP tool documentation for saving competitor intelligence reports
+
+## Available Commands
+
+- `notion-create-pages` - Create new pages in the SEO Audit Log database
+- `notion-update-page` - Update existing audit entries
+- `notion-query-database-view` - Query existing reports
+- `notion-search` - Search across Notion workspace
+
+## Configuration
+
+- Database ID: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- All reports saved with Category: "Competitor Intelligence"
+- Audit ID format: COMP-YYYYMMDD-NNN
+
+## Examples
+
+```
+# Create competitor intelligence report
+mcp__notion__notion-create-pages(
+  database_id="2c8581e5-8a1e-8035-880b-e38cefc2f3ef",
+  properties={
+    "Issue": {"title": [{"text": {"content": "경쟁사 분석 보고서 - 2025-01-15"}}]},
+    "Site": {"url": "https://example.com"},
+    "Category": {"select": {"name": "Competitor Intelligence"}},
+    "Priority": {"select": {"name": "High"}},
+    "Found Date": {"date": {"start": "2025-01-15"}},
+    "Audit ID": {"rich_text": [{"text": {"content": "COMP-20250115-001"}}]}
+  }
+)
+```
diff --git a/custom-skills/31-seo-competitor-intel/desktop/tools/websearch.md b/custom-skills/31-seo-competitor-intel/desktop/tools/websearch.md
new file mode 100644
index 0000000..e5aa101
--- /dev/null
+++ b/custom-skills/31-seo-competitor-intel/desktop/tools/websearch.md
@@ -0,0 +1,28 @@
+# WebSearch
+
+> Web search tool documentation for competitor intelligence
+
+## Available Commands
+
+- `WebSearch` - Search the web for competitor information, market data, and Korean market presence
+
+## Configuration
+
+- No additional configuration required
+- Available as built-in Claude tool
+
+## Examples
+
+```
+# Check Naver Blog presence
+WebSearch(query="site:blog.naver.com competitor-domain.com")
+
+# Check Naver Cafe presence
+WebSearch(query="site:cafe.naver.com competitor-domain.com")
+
+# Research competitor market presence
+WebSearch(query="competitor-domain.com market share industry analysis")
+
+# Korean market competitor discovery
+WebSearch(query="경쟁사 분석 industry-keyword")
+```
diff --git a/custom-skills/32-seo-crawl-budget/code/CLAUDE.md b/custom-skills/32-seo-crawl-budget/code/CLAUDE.md
new file mode 100644
index 0000000..1f540e4
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/code/CLAUDE.md
@@ -0,0 +1,178 @@
+# CLAUDE.md
+
+## Overview
+
+Crawl budget optimization tool for analyzing server access logs and identifying crawl budget waste. Parses Apache/Nginx/CloudFront access logs, identifies search engine bots (Googlebot, Yeti/Naver, Bingbot, Daumoa/Kakao), profiles per-bot crawl behavior, detects crawl waste (parameter URLs, low-value pages, redirect chains), identifies orphan pages, and generates crawl efficiency recommendations. Uses streaming parser for large log files.
+
+## Quick Start
+
+```bash
+pip install -r scripts/requirements.txt
+
+# Parse access logs
+python scripts/log_parser.py --log-file /var/log/nginx/access.log --json
+
+# Crawl budget analysis
+python scripts/crawl_budget_analyzer.py --log-file /var/log/nginx/access.log --sitemap https://example.com/sitemap.xml --json
+```
+
+## Scripts
+
+| Script | Purpose | Key Output |
+|--------|---------|------------|
+| `log_parser.py` | Parse server access logs, identify bots, extract crawl data | Bot identification, request patterns, status codes |
+| `crawl_budget_analyzer.py` | Analyze crawl budget efficiency and generate recommendations | Waste identification, orphan pages, optimization plan |
+| `base_client.py` | Shared utilities | RateLimiter, ConfigManager, BaseAsyncClient |
+
+## Log Parser
+
+```bash
+# Parse Nginx combined log format
+python scripts/log_parser.py --log-file /var/log/nginx/access.log --json
+
+# Parse Apache combined log format
+python scripts/log_parser.py --log-file /var/log/apache2/access.log --format apache --json
+
+# Parse CloudFront log
+python scripts/log_parser.py --log-file cloudfront-log.gz --format cloudfront --json
+
+# Filter by specific bot
+python scripts/log_parser.py --log-file access.log --bot googlebot --json
+
+# Parse gzipped logs
+python scripts/log_parser.py --log-file access.log.gz --json
+
+# Process large files in streaming mode
+python scripts/log_parser.py --log-file access.log --streaming --json
+```
+
+**Capabilities**:
+- Support for common log formats:
+  - Nginx combined format
+  - Apache combined format
+  - CloudFront format
+  - Custom format via regex
+- Bot identification by User-Agent:
+  - Googlebot (and variants: Googlebot-Image, Googlebot-News, Googlebot-Video, AdsBot-Google)
+  - Yeti (Naver's crawler)
+  - Bingbot
+  - Daumoa (Kakao/Daum crawler)
+  - Other bots (Applebot, DuckDuckBot, Baiduspider, etc.)
+- Request data extraction (timestamp, IP, URL, status code, response size, user-agent, referer)
+- Streaming parser for files >1GB
+- Gzip/bzip2 compressed log support
+- Date range filtering
+
+## Crawl Budget Analyzer
+
+```bash
+# Full crawl budget analysis
+python scripts/crawl_budget_analyzer.py --log-file access.log --sitemap https://example.com/sitemap.xml --json
+
+# Waste identification only
+python scripts/crawl_budget_analyzer.py --log-file access.log --scope waste --json
+
+# Orphan page detection
+python scripts/crawl_budget_analyzer.py --log-file access.log --sitemap https://example.com/sitemap.xml --scope orphans --json
+
+# Per-bot profiling
+python scripts/crawl_budget_analyzer.py --log-file access.log --scope bots --json
+
+# With Ahrefs page history comparison
+python scripts/crawl_budget_analyzer.py --log-file access.log --url https://example.com --ahrefs --json
+```
+
+**Capabilities**:
+- Crawl budget waste identification:
+  - Parameter URLs consuming crawl budget (?sort=, ?filter=, ?page=, ?utm_*)
+  - Low-value pages (thin content, noindex pages being crawled)
+  - Redirect chains consuming multiple crawls
+  - Soft 404 pages (200 status but error content)
+  - Duplicate URLs (www/non-www, http/https, trailing slash variants)
+- Per-bot behavior profiling:
+  - Crawl frequency (requests/day, requests/hour)
+  - Crawl depth distribution
+  - Status code distribution per bot
+  - Most crawled URLs per bot
+  - Crawl pattern analysis (time of day, days of week)
+- Orphan page detection:
+  - Pages in sitemap but never crawled by bots
+  - Pages crawled but not in sitemap
+  - Crawled pages with no internal links
+- Crawl efficiency recommendations:
+  - robots.txt optimization suggestions
+  - URL parameter handling recommendations
+  - Noindex/nofollow suggestions for low-value pages
+  - Redirect chain resolution priorities
+  - Internal linking improvements for orphan pages
+
+## Data Sources
+
+| Source | Purpose |
+|--------|---------|
+| Server access logs | Primary crawl data |
+| XML sitemap | Reference for expected crawlable pages |
+| Ahrefs `site-explorer-pages-history` | Compare indexed pages with crawled pages |
+
+## Output Format
+
+```json
+{
+  "log_file": "access.log",
+  "analysis_period": {"from": "2025-01-01", "to": "2025-01-31"},
+  "total_bot_requests": 150000,
+  "bots": {
+    "googlebot": {
+      "requests": 80000,
+      "unique_urls": 12000,
+      "avg_requests_per_day": 2580,
+      "status_distribution": {"200": 70000, "301": 5000, "404": 3000, "500": 2000},
+      "top_crawled_urls": [...]
+    },
+    "yeti": {"requests": 35000, ...},
+    "bingbot": {"requests": 20000, ...},
+    "daumoa": {"requests": 15000, ...}
+  },
+  "waste": {
+    "parameter_urls": {"count": 5000, "pct_of_crawls": 3.3},
+    "redirect_chains": {"count": 2000, "pct_of_crawls": 1.3},
+    "soft_404s": {"count": 1500, "pct_of_crawls": 1.0},
+    "total_waste_pct": 8.5
+  },
+  "orphan_pages": {
+    "in_sitemap_not_crawled": [...],
+    "crawled_not_in_sitemap": [...]
+  },
+  "recommendations": [...],
+  "efficiency_score": 72,
+  "timestamp": "2025-01-01T00:00:00"
+}
+```
+
+## Notion Output (Required)
+
+**IMPORTANT**: All audit reports MUST be saved to the OurDigital SEO Audit Log database.
+
+### Database Configuration
+
+| Field | Value |
+|-------|-------|
+| Database ID | `2c8581e5-8a1e-8035-880b-e38cefc2f3ef` |
+| URL | https://www.notion.so/dintelligence/2c8581e58a1e8035880be38cefc2f3ef |
+
+### Required Properties
+
+| Property | Type | Description |
+|----------|------|-------------|
+| Issue | Title | Report title (Korean + date) |
+| Site | URL | Audited website URL |
+| Category | Select | Crawl Budget |
+| Priority | Select | Based on waste percentage |
+| Found Date | Date | Audit date (YYYY-MM-DD) |
+| Audit ID | Rich Text | Format: CRAWL-YYYYMMDD-NNN |
+
+### Language Guidelines
+
+- Report content in Korean (한국어)
+- Keep technical English terms as-is (e.g., Crawl Budget, Googlebot, robots.txt)
+- URLs and code remain unchanged
diff --git a/custom-skills/32-seo-crawl-budget/code/scripts/base_client.py b/custom-skills/32-seo-crawl-budget/code/scripts/base_client.py
new file mode 100644
index 0000000..ac5715b
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/code/scripts/base_client.py
@@ -0,0 +1,207 @@
+"""
+Base Client - Shared async client utilities
+===========================================
+Purpose: Rate-limited async operations for API clients
+Python: 3.10+
+"""
+
+import asyncio
+import logging
+import os
+from asyncio import Semaphore
+from datetime import datetime
+from typing import Any, Callable, TypeVar
+
+from dotenv import load_dotenv
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+)
+
+# Load environment variables
+load_dotenv()
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+
+T = TypeVar("T")
+
+
+class RateLimiter:
+    """Rate limiter using token bucket algorithm."""
+
+    def __init__(self, rate: float, per: float = 1.0):
+        """
+        Initialize rate limiter.
+
+        Args:
+            rate: Number of requests allowed
+            per: Time period in seconds (default: 1 second)
+        """
+        self.rate = rate
+        self.per = per
+        self.tokens = rate
+        self.last_update = datetime.now()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self) -> None:
+        """Acquire a token, waiting if necessary."""
+        async with self._lock:
+            now = datetime.now()
+            elapsed = (now - self.last_update).total_seconds()
+            self.tokens = min(self.rate, self.tokens + elapsed * (self.rate / self.per))
+            self.last_update = now
+
+            if self.tokens < 1:
+                wait_time = (1 - self.tokens) * (self.per / self.rate)
+                await asyncio.sleep(wait_time)
+                self.tokens = 0
+            else:
+                self.tokens -= 1
+
+
+class BaseAsyncClient:
+    """Base class for async API clients with rate limiting."""
+
+    def __init__(
+        self,
+        max_concurrent: int = 5,
+        requests_per_second: float = 3.0,
+        logger: logging.Logger | None = None,
+    ):
+        """
+        Initialize base client.
+
+        Args:
+            max_concurrent: Maximum concurrent requests
+            requests_per_second: Rate limit
+            logger: Logger instance
+        """
+        self.semaphore = Semaphore(max_concurrent)
+        self.rate_limiter = RateLimiter(requests_per_second)
+        self.logger = logger or logging.getLogger(self.__class__.__name__)
+        self.stats = {
+            "requests": 0,
+            "success": 0,
+            "errors": 0,
+            "retries": 0,
+        }
+
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=2, max=10),
+        retry=retry_if_exception_type(Exception),
+    )
+    async def _rate_limited_request(
+        self,
+        coro: Callable[[], Any],
+    ) -> Any:
+        """Execute a request with rate limiting and retry."""
+        async with self.semaphore:
+            await self.rate_limiter.acquire()
+            self.stats["requests"] += 1
+            try:
+                result = await coro()
+                self.stats["success"] += 1
+                return result
+            except Exception as e:
+                self.stats["errors"] += 1
+                self.logger.error(f"Request failed: {e}")
+                raise
+
+    async def batch_requests(
+        self,
+        requests: list[Callable[[], Any]],
+        desc: str = "Processing",
+    ) -> list[Any]:
+        """Execute multiple requests concurrently."""
+        try:
+            from tqdm.asyncio import tqdm
+            has_tqdm = True
+        except ImportError:
+            has_tqdm = False
+
+        async def execute(req: Callable) -> Any:
+            try:
+                return await self._rate_limited_request(req)
+            except Exception as e:
+                return {"error": str(e)}
+
+        tasks = [execute(req) for req in requests]
+
+        if has_tqdm:
+            results = []
+            for coro in tqdm.as_completed(tasks, total=len(tasks), desc=desc):
+                result = await coro
+                results.append(result)
+            return results
+        else:
+            return await asyncio.gather(*tasks, return_exceptions=True)
+
+    def print_stats(self) -> None:
+        """Print request statistics."""
+        self.logger.info("=" * 40)
+        self.logger.info("Request Statistics:")
+        self.logger.info(f"  Total Requests: {self.stats['requests']}")
+        self.logger.info(f"  Successful: {self.stats['success']}")
+        self.logger.info(f"  Errors: {self.stats['errors']}")
+        self.logger.info("=" * 40)
+
+
+class ConfigManager:
+    """Manage API configuration and credentials."""
+
+    def __init__(self):
+        load_dotenv()
+
+    @property
+    def google_credentials_path(self) -> str | None:
+        """Get Google service account credentials path."""
+        # Prefer SEO-specific credentials, fallback to general credentials
+        seo_creds = os.path.expanduser("~/.credential/ourdigital-seo-agent.json")
+        if os.path.exists(seo_creds):
+            return seo_creds
+        return os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
+
+    @property
+    def pagespeed_api_key(self) -> str | None:
+        """Get PageSpeed Insights API key."""
+        return os.getenv("PAGESPEED_API_KEY")
+
+    @property
+    def custom_search_api_key(self) -> str | None:
+        """Get Custom Search API key."""
+        return os.getenv("CUSTOM_SEARCH_API_KEY")
+
+    @property
+    def custom_search_engine_id(self) -> str | None:
+        """Get Custom Search Engine ID."""
+        return os.getenv("CUSTOM_SEARCH_ENGINE_ID")
+
+    @property
+    def notion_token(self) -> str | None:
+        """Get Notion API token."""
+        return os.getenv("NOTION_TOKEN") or os.getenv("NOTION_API_KEY")
+
+    def validate_google_credentials(self) -> bool:
+        """Validate Google credentials are configured."""
+        creds_path = self.google_credentials_path
+        if not creds_path:
+            return False
+        return os.path.exists(creds_path)
+
+    def get_required(self, key: str) -> str:
+        """Get required environment variable or raise error."""
+        value = os.getenv(key)
+        if not value:
+            raise ValueError(f"Missing required environment variable: {key}")
+        return value
+
+
+# Singleton config instance
+config = ConfigManager()
diff --git a/custom-skills/32-seo-crawl-budget/code/scripts/crawl_budget_analyzer.py b/custom-skills/32-seo-crawl-budget/code/scripts/crawl_budget_analyzer.py
new file mode 100644
index 0000000..a0b10dc
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/code/scripts/crawl_budget_analyzer.py
@@ -0,0 +1,805 @@
+"""
+Crawl Budget Analyzer - Identify crawl waste and generate recommendations
+=========================================================================
+Purpose: Analyze server access logs for crawl budget efficiency, detect waste
+         (parameter URLs, redirect chains, soft 404s, duplicates), find orphan
+         pages, profile per-bot behavior, and produce prioritized recommendations.
+Python: 3.10+
+"""
+
+import argparse
+import json
+import logging
+import re
+import sys
+from collections import Counter, defaultdict
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from pathlib import Path
+from typing import Any
+from urllib.parse import parse_qs, urlparse
+
+import requests
+from bs4 import BeautifulSoup
+
+from log_parser import BotIdentification, LogEntry, LogParser
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Constants
+# ---------------------------------------------------------------------------
+
+WASTE_PARAMS = {"sort", "filter", "order", "orderby", "dir", "direction"}
+TRACKING_PARAMS_RE = re.compile(r"^utm_", re.IGNORECASE)
+PAGINATION_PARAM = "page"
+HIGH_PAGE_THRESHOLD = 5
+SOFT_404_MAX_SIZE = 1024  # bytes - pages smaller than this may be soft 404s
+REDIRECT_STATUSES = {301, 302, 303, 307, 308}
+TOP_N_URLS = 50
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class CrawlWaste:
+    """A category of crawl budget waste."""
+    waste_type: str
+    urls: list[str]
+    count: int
+    pct_of_total: float
+    recommendation: str
+
+    def to_dict(self) -> dict:
+        return {
+            "waste_type": self.waste_type,
+            "count": self.count,
+            "pct_of_total": round(self.pct_of_total, 2),
+            "recommendation": self.recommendation,
+            "sample_urls": self.urls[:20],
+        }
+
+
+@dataclass
+class OrphanPage:
+    """A page that is either in the sitemap but uncrawled, or crawled but not in sitemap."""
+    url: str
+    in_sitemap: bool
+    crawled: bool
+    last_crawl_date: str | None = None
+
+    def to_dict(self) -> dict:
+        return asdict(self)
+
+
+@dataclass
+class BotProfile:
+    """Per-bot crawl behavior profile."""
+    name: str
+    total_requests: int = 0
+    requests_per_day: float = 0.0
+    crawl_depth_distribution: dict[int, int] = field(default_factory=dict)
+    peak_hours: list[int] = field(default_factory=list)
+    status_breakdown: dict[str, int] = field(default_factory=dict)
+    top_crawled_urls: list[tuple[str, int]] = field(default_factory=list)
+    unique_urls: int = 0
+    days_active: int = 0
+
+    def to_dict(self) -> dict:
+        return {
+            "name": self.name,
+            "total_requests": self.total_requests,
+            "requests_per_day": round(self.requests_per_day, 1),
+            "crawl_depth_distribution": self.crawl_depth_distribution,
+            "peak_hours": self.peak_hours,
+            "status_breakdown": self.status_breakdown,
+            "top_crawled_urls": [{"url": u, "count": c} for u, c in self.top_crawled_urls],
+            "unique_urls": self.unique_urls,
+            "days_active": self.days_active,
+        }
+
+
+@dataclass
+class CrawlRecommendation:
+    """A single optimization recommendation."""
+    category: str
+    priority: str  # critical, high, medium, low
+    action: str
+    impact: str
+    details: str
+
+    def to_dict(self) -> dict:
+        return asdict(self)
+
+
+@dataclass
+class CrawlBudgetResult:
+    """Complete crawl budget analysis result."""
+    log_file: str
+    analysis_period: dict[str, str]
+    total_bot_requests: int
+    bots: dict[str, BotProfile]
+    waste: list[CrawlWaste]
+    total_waste_pct: float
+    orphan_pages: dict[str, list[OrphanPage]]
+    recommendations: list[CrawlRecommendation]
+    efficiency_score: int
+    timestamp: str
+
+    def to_dict(self) -> dict:
+        return {
+            "log_file": self.log_file,
+            "analysis_period": self.analysis_period,
+            "total_bot_requests": self.total_bot_requests,
+            "bots": {n: p.to_dict() for n, p in self.bots.items()},
+            "waste": {w.waste_type: w.to_dict() for w in self.waste},
+            "total_waste_pct": round(self.total_waste_pct, 2),
+            "orphan_pages": {
+                k: [o.to_dict() for o in v]
+                for k, v in self.orphan_pages.items()
+            },
+            "recommendations": [r.to_dict() for r in self.recommendations],
+            "efficiency_score": self.efficiency_score,
+            "timestamp": self.timestamp,
+        }
+
+
+# ---------------------------------------------------------------------------
+# CrawlBudgetAnalyzer
+# ---------------------------------------------------------------------------
+
+class CrawlBudgetAnalyzer:
+    """Analyze crawl budget efficiency from server access logs."""
+
+    def __init__(
+        self,
+        log_file: str,
+        sitemap_url: str | None = None,
+        target_url: str | None = None,
+    ):
+        self.log_file = log_file
+        self.sitemap_url = sitemap_url
+        self.target_url = target_url
+        self._bot_entries: list[tuple[LogEntry, BotIdentification]] = []
+        self._sitemap_urls: set[str] = set()
+
+    # -- data loading ---------------------------------------------------------
+
+    def load_log_data(self, log_file: str) -> list[tuple[LogEntry, BotIdentification]]:
+        """Use LogParser to load all bot requests from the log file."""
+        parser = LogParser(log_file=log_file, fmt="auto")
+        entries = parser.parse()
+        logger.info(f"Loaded {len(entries):,} bot entries from {log_file}")
+        self._bot_entries = entries
+        return entries
+
+    def load_sitemap_urls(self, sitemap_url: str) -> set[str]:
+        """Fetch and parse an XML sitemap, returning the set of URLs."""
+        urls: set[str] = set()
+        try:
+            resp = requests.get(sitemap_url, timeout=30, headers={
+                "User-Agent": "CrawlBudgetAnalyzer/1.0",
+            })
+            resp.raise_for_status()
+            soup = BeautifulSoup(resp.content, "lxml-xml")
+
+            # Handle sitemap index
+            sitemap_tags = soup.find_all("sitemap")
+            if sitemap_tags:
+                for st in sitemap_tags:
+                    loc = st.find("loc")
+                    if loc and loc.text:
+                        child_urls = self._fetch_sitemap_child(loc.text.strip())
+                        urls.update(child_urls)
+            else:
+                for url_tag in soup.find_all("url"):
+                    loc = url_tag.find("loc")
+                    if loc and loc.text:
+                        urls.add(self._normalize_url(loc.text.strip()))
+
+            logger.info(f"Loaded {len(urls):,} URLs from sitemap: {sitemap_url}")
+        except Exception as e:
+            logger.error(f"Failed to load sitemap {sitemap_url}: {e}")
+
+        self._sitemap_urls = urls
+        return urls
+
+    def _fetch_sitemap_child(self, url: str) -> set[str]:
+        """Fetch a child sitemap from a sitemap index."""
+        urls: set[str] = set()
+        try:
+            resp = requests.get(url, timeout=30, headers={
+                "User-Agent": "CrawlBudgetAnalyzer/1.0",
+            })
+            resp.raise_for_status()
+            soup = BeautifulSoup(resp.content, "lxml-xml")
+            for url_tag in soup.find_all("url"):
+                loc = url_tag.find("loc")
+                if loc and loc.text:
+                    urls.add(self._normalize_url(loc.text.strip()))
+        except Exception as e:
+            logger.warning(f"Failed to fetch child sitemap {url}: {e}")
+        return urls
+
+    @staticmethod
+    def _normalize_url(url: str) -> str:
+        """Normalize a URL by removing trailing slash and lowercasing the scheme/host."""
+        parsed = urlparse(url)
+        path = parsed.path.rstrip("/") or "/"
+        return f"{parsed.scheme}://{parsed.netloc}{path}"
+
+    # -- waste identification -------------------------------------------------
+
+    def identify_parameter_waste(
+        self,
+        bot_requests: list[tuple[LogEntry, BotIdentification]],
+    ) -> CrawlWaste:
+        """Find URLs with unnecessary query parameters wasting crawl budget."""
+        waste_urls: list[str] = []
+        for entry, _ in bot_requests:
+            parsed = urlparse(entry.url)
+            if not parsed.query:
+                continue
+            params = parse_qs(parsed.query)
+            param_keys = {k.lower() for k in params}
+            # Check for waste parameters
+            has_waste = bool(param_keys & WASTE_PARAMS)
+            # Check for tracking parameters
+            has_tracking = any(TRACKING_PARAMS_RE.match(k) for k in param_keys)
+            # Check for deep pagination
+            page_val = params.get(PAGINATION_PARAM, params.get("p", [None]))
+            has_deep_page = False
+            if page_val and page_val[0]:
+                try:
+                    if int(page_val[0]) > HIGH_PAGE_THRESHOLD:
+                        has_deep_page = True
+                except (ValueError, TypeError):
+                    pass
+            if has_waste or has_tracking or has_deep_page:
+                waste_urls.append(entry.url)
+
+        total = len(bot_requests)
+        count = len(waste_urls)
+        pct = (count / total * 100) if total else 0.0
+        return CrawlWaste(
+            waste_type="parameter_urls",
+            urls=list(set(waste_urls)),
+            count=count,
+            pct_of_total=pct,
+            recommendation=(
+                "robots.txt에 불필요한 parameter URL 패턴을 Disallow로 추가하거나, "
+                "Google Search Console의 URL Parameters 설정을 활용하세요. "
+                "UTM 파라미터가 포함된 URL은 canonical 태그로 처리하세요."
+            ),
+        )
+
+    def identify_redirect_chains(
+        self,
+        bot_requests: list[tuple[LogEntry, BotIdentification]],
+    ) -> CrawlWaste:
+        """Find URLs that repeatedly return redirect status codes."""
+        redirect_urls: list[str] = []
+        redirect_counter: Counter = Counter()
+        for entry, _ in bot_requests:
+            if entry.status_code in REDIRECT_STATUSES:
+                redirect_counter[entry.url] += 1
+                redirect_urls.append(entry.url)
+
+        # URLs redirected more than once are chain candidates
+        chain_urls = [url for url, cnt in redirect_counter.items() if cnt >= 2]
+        total = len(bot_requests)
+        count = len(redirect_urls)
+        pct = (count / total * 100) if total else 0.0
+        return CrawlWaste(
+            waste_type="redirect_chains",
+            urls=chain_urls,
+            count=count,
+            pct_of_total=pct,
+            recommendation=(
+                "301/302 리다이렉트가 반복적으로 크롤링되고 있습니다. "
+                "내부 링크를 최종 목적지 URL로 직접 업데이트하고, "
+                "리다이렉트 체인을 단일 리다이렉트로 단축하세요."
+            ),
+        )
+
+    def identify_soft_404s(
+        self,
+        bot_requests: list[tuple[LogEntry, BotIdentification]],
+    ) -> CrawlWaste:
+        """Find 200-status pages with suspiciously small response sizes."""
+        soft_404_urls: list[str] = []
+        for entry, _ in bot_requests:
+            if entry.status_code == 200 and entry.response_size < SOFT_404_MAX_SIZE:
+                if entry.response_size > 0:
+                    soft_404_urls.append(entry.url)
+
+        total = len(bot_requests)
+        count = len(soft_404_urls)
+        pct = (count / total * 100) if total else 0.0
+        return CrawlWaste(
+            waste_type="soft_404s",
+            urls=list(set(soft_404_urls)),
+            count=count,
+            pct_of_total=pct,
+            recommendation=(
+                "200 상태 코드를 반환하지만 콘텐츠가 거의 없는 Soft 404 페이지입니다. "
+                "실제 404 상태 코드를 반환하거나, 해당 페이지에 noindex 태그를 추가하세요."
+            ),
+        )
+
+    def identify_duplicate_crawls(
+        self,
+        bot_requests: list[tuple[LogEntry, BotIdentification]],
+    ) -> CrawlWaste:
+        """Find duplicate URL variants: www/non-www, trailing slash, etc."""
+        url_variants: dict[str, set[str]] = defaultdict(set)
+        for entry, _ in bot_requests:
+            parsed = urlparse(entry.url)
+            # Normalize: strip www, strip trailing slash, lowercase
+            host = parsed.netloc.lower().lstrip("www.")
+            path = parsed.path.rstrip("/") or "/"
+            canonical = f"{host}{path}"
+            full_url = f"{parsed.scheme}://{parsed.netloc}{parsed.path}"
+            url_variants[canonical].add(full_url)
+
+        # Identify canonicals with multiple variants
+        duplicate_urls: list[str] = []
+        for canonical, variants in url_variants.items():
+            if len(variants) > 1:
+                duplicate_urls.extend(variants)
+
+        total = len(bot_requests)
+        # Count how many requests hit duplicate variant URLs
+        dup_set = set(duplicate_urls)
+        dup_request_count = sum(1 for e, _ in bot_requests if f"{urlparse(e.url).scheme}://{urlparse(e.url).netloc}{urlparse(e.url).path}" in dup_set)
+        pct = (dup_request_count / total * 100) if total else 0.0
+        return CrawlWaste(
+            waste_type="duplicate_urls",
+            urls=duplicate_urls[:TOP_N_URLS],
+            count=dup_request_count,
+            pct_of_total=pct,
+            recommendation=(
+                "www/non-www, trailing slash 유무 등 중복 URL 변형이 크롤링되고 있습니다. "
+                "301 리다이렉트로 canonical URL로 통합하고, "
+                "rel=canonical 태그를 정확히 설정하세요."
+            ),
+        )
+
+    # -- bot profiling --------------------------------------------------------
+
+    def profile_bots(
+        self,
+        bot_requests: list[tuple[LogEntry, BotIdentification]],
+    ) -> dict[str, BotProfile]:
+        """Generate per-bot behavior profiles."""
+        bot_data: dict[str, dict] = defaultdict(lambda: {
+            "urls": Counter(),
+            "statuses": Counter(),
+            "hours": Counter(),
+            "days": set(),
+            "depths": Counter(),
+            "count": 0,
+        })
+
+        for entry, bot in bot_requests:
+            bd = bot_data[bot.name]
+            bd["count"] += 1
+            bd["urls"][entry.url] += 1
+            bd["statuses"][str(entry.status_code)] += 1
+            # URL depth = number of path segments
+            depth = len([s for s in urlparse(entry.url).path.split("/") if s])
+            bd["depths"][depth] += 1
+            if entry.timestamp:
+                bd["hours"][entry.timestamp.hour] += 1
+                bd["days"].add(entry.timestamp.strftime("%Y-%m-%d"))
+
+        profiles: dict[str, BotProfile] = {}
+        for name, bd in bot_data.items():
+            days_active = len(bd["days"]) or 1
+            rpd = bd["count"] / days_active
+            # Top 3 peak hours
+            top_hours = sorted(bd["hours"].items(), key=lambda x: -x[1])[:3]
+            peak = [h for h, _ in top_hours]
+            profiles[name] = BotProfile(
+                name=name,
+                total_requests=bd["count"],
+                requests_per_day=rpd,
+                crawl_depth_distribution=dict(sorted(bd["depths"].items())),
+                peak_hours=peak,
+                status_breakdown=dict(bd["statuses"]),
+                top_crawled_urls=bd["urls"].most_common(TOP_N_URLS),
+                unique_urls=len(bd["urls"]),
+                days_active=days_active,
+            )
+        return profiles
+
+    # -- orphan detection -----------------------------------------------------
+
+    def detect_orphan_pages(
+        self,
+        crawled_urls: set[str],
+        sitemap_urls: set[str],
+    ) -> dict[str, list[OrphanPage]]:
+        """Compare crawled URLs with sitemap URLs to find orphans."""
+        in_sitemap_not_crawled = sitemap_urls - crawled_urls
+        crawled_not_in_sitemap = crawled_urls - sitemap_urls
+
+        return {
+            "in_sitemap_not_crawled": [
+                OrphanPage(url=u, in_sitemap=True, crawled=False)
+                for u in sorted(in_sitemap_not_crawled)
+            ],
+            "crawled_not_in_sitemap": [
+                OrphanPage(url=u, in_sitemap=False, crawled=True)
+                for u in sorted(crawled_not_in_sitemap)
+            ],
+        }
+
+    # -- efficiency score -----------------------------------------------------
+
+    @staticmethod
+    def calculate_efficiency_score(total_waste_pct: float) -> int:
+        """Calculate crawl efficiency score: 100 - waste%, capped at [0, 100]."""
+        score = int(100 - total_waste_pct)
+        return max(0, min(100, score))
+
+    # -- recommendations ------------------------------------------------------
+
+    def generate_recommendations(
+        self,
+        waste: list[CrawlWaste],
+        orphans: dict[str, list[OrphanPage]],
+        bot_profiles: dict[str, BotProfile],
+    ) -> list[CrawlRecommendation]:
+        """Generate prioritized crawl budget optimization recommendations."""
+        recs: list[CrawlRecommendation] = []
+
+        # Waste-based recommendations
+        for w in waste:
+            if w.pct_of_total > 5.0:
+                priority = "critical"
+            elif w.pct_of_total > 2.0:
+                priority = "high"
+            elif w.pct_of_total > 0.5:
+                priority = "medium"
+            else:
+                priority = "low"
+
+            if w.waste_type == "parameter_urls" and w.count > 0:
+                recs.append(CrawlRecommendation(
+                    category="URL Parameters",
+                    priority=priority,
+                    action="robots.txt에 parameter URL 패턴 Disallow 규칙 추가",
+                    impact=f"크롤 요청 {w.pct_of_total:.1f}% 절감 가능",
+                    details=(
+                        f"총 {w.count:,}건의 parameter URL이 크롤링되었습니다. "
+                        f"sort, filter, utm_* 등 불필요한 파라미터를 차단하세요."
+                    ),
+                ))
+            elif w.waste_type == "redirect_chains" and w.count > 0:
+                recs.append(CrawlRecommendation(
+                    category="Redirect Chains",
+                    priority=priority,
+                    action="리다이렉트 체인을 단축하고 내부 링크 업데이트",
+                    impact=f"크롤 요청 {w.pct_of_total:.1f}% 절감 가능",
+                    details=(
+                        f"총 {w.count:,}건의 리다이렉트 요청이 발생했습니다. "
+                        f"내부 링크를 최종 URL로 직접 연결하세요."
+                    ),
+                ))
+            elif w.waste_type == "soft_404s" and w.count > 0:
+                recs.append(CrawlRecommendation(
+                    category="Soft 404s",
+                    priority=priority,
+                    action="Soft 404 페이지에 적절한 HTTP 상태 코드 또는 noindex 적용",
+                    impact=f"크롤 요청 {w.pct_of_total:.1f}% 절감 가능",
+                    details=(
+                        f"총 {w.count:,}건의 Soft 404가 감지되었습니다. "
+                        f"적절한 404 응답 또는 noindex meta 태그를 설정하세요."
+                    ),
+                ))
+            elif w.waste_type == "duplicate_urls" and w.count > 0:
+                recs.append(CrawlRecommendation(
+                    category="Duplicate URLs",
+                    priority=priority,
+                    action="URL 정규화 및 canonical 태그 설정",
+                    impact=f"크롤 요청 {w.pct_of_total:.1f}% 절감 가능",
+                    details=(
+                        f"총 {w.count:,}건의 중복 URL 변형이 크롤링되었습니다. "
+                        f"www/non-www, trailing slash 통합을 진행하세요."
+                    ),
+                ))
+
+        # Orphan page recommendations
+        not_crawled = orphans.get("in_sitemap_not_crawled", [])
+        not_in_sitemap = orphans.get("crawled_not_in_sitemap", [])
+
+        if len(not_crawled) > 0:
+            pct = len(not_crawled) / max(len(self._sitemap_urls), 1) * 100
+            priority = "critical" if pct > 30 else "high" if pct > 10 else "medium"
+            recs.append(CrawlRecommendation(
+                category="Orphan Pages (Uncrawled)",
+                priority=priority,
+                action="사이트맵에 있으나 크롤링되지 않은 페이지의 내부 링크 강화",
+                impact=f"사이트맵 URL의 {pct:.1f}%가 미크롤 상태",
+                details=(
+                    f"총 {len(not_crawled):,}개 URL이 사이트맵에 있지만 "
+                    f"봇이 크롤링하지 않았습니다. 내부 링크를 추가하세요."
+                ),
+            ))
+
+        if len(not_in_sitemap) > 0:
+            recs.append(CrawlRecommendation(
+                category="Orphan Pages (Unlisted)",
+                priority="medium",
+                action="크롤링되었으나 사이트맵에 없는 페이지를 사이트맵에 추가 또는 차단",
+                impact=f"{len(not_in_sitemap):,}개 URL이 사이트맵에 미등록",
+                details=(
+                    f"봇이 크롤링한 {len(not_in_sitemap):,}개 URL이 "
+                    f"사이트맵에 포함되어 있지 않습니다. 유효한 페이지는 "
+                    f"사이트맵에 추가하고, 불필요한 페이지는 robots.txt로 차단하세요."
+                ),
+            ))
+
+        # Bot-specific recommendations
+        for name, profile in bot_profiles.items():
+            error_count = sum(
+                v for k, v in profile.status_breakdown.items()
+                if k.startswith("4") or k.startswith("5")
+            )
+            error_pct = (error_count / profile.total_requests * 100) if profile.total_requests else 0
+            if error_pct > 10:
+                recs.append(CrawlRecommendation(
+                    category=f"Bot Errors ({name})",
+                    priority="high" if error_pct > 20 else "medium",
+                    action=f"{name}의 4xx/5xx 오류율 {error_pct:.1f}% 개선 필요",
+                    impact=f"{name} 크롤 예산의 {error_pct:.1f}%가 오류에 소비",
+                    details=(
+                        f"{name}이(가) {error_count:,}건의 오류 응답을 받았습니다. "
+                        f"깨진 링크를 수정하고 서버 안정성을 개선하세요."
+                    ),
+                ))
+
+        # Sort by priority
+        priority_order = {"critical": 0, "high": 1, "medium": 2, "low": 3}
+        recs.sort(key=lambda r: priority_order.get(r.priority, 4))
+        return recs
+
+    # -- orchestrator ---------------------------------------------------------
+
+    def analyze(self, scope: str = "all") -> CrawlBudgetResult:
+        """Orchestrate the full crawl budget analysis."""
+        # Load log data
+        entries = self.load_log_data(self.log_file)
+        if not entries:
+            logger.warning("No bot entries found in log file.")
+
+        # Load sitemap if provided
+        if self.sitemap_url:
+            self.load_sitemap_urls(self.sitemap_url)
+
+        # Profile bots
+        bot_profiles: dict[str, BotProfile] = {}
+        if scope in ("all", "bots"):
+            bot_profiles = self.profile_bots(entries)
+
+        # Identify waste
+        waste: list[CrawlWaste] = []
+        if scope in ("all", "waste"):
+            waste.append(self.identify_parameter_waste(entries))
+            waste.append(self.identify_redirect_chains(entries))
+            waste.append(self.identify_soft_404s(entries))
+            waste.append(self.identify_duplicate_crawls(entries))
+
+        total_waste_pct = sum(w.pct_of_total for w in waste)
+
+        # Detect orphan pages
+        orphans: dict[str, list[OrphanPage]] = {
+            "in_sitemap_not_crawled": [],
+            "crawled_not_in_sitemap": [],
+        }
+        if scope in ("all", "orphans") and self._sitemap_urls:
+            crawled_urls: set[str] = set()
+            for entry, _ in entries:
+                # Build full URL from path for comparison
+                if self.target_url:
+                    parsed_target = urlparse(self.target_url)
+                    full = f"{parsed_target.scheme}://{parsed_target.netloc}{entry.url}"
+                    crawled_urls.add(self._normalize_url(full))
+                else:
+                    crawled_urls.add(entry.url)
+            orphans = self.detect_orphan_pages(crawled_urls, self._sitemap_urls)
+
+        # Efficiency score
+        efficiency_score = self.calculate_efficiency_score(total_waste_pct)
+
+        # Recommendations
+        recommendations = self.generate_recommendations(waste, orphans, bot_profiles)
+
+        # Date range from entries
+        timestamps = [e.timestamp for e, _ in entries if e.timestamp]
+        analysis_period = {}
+        if timestamps:
+            analysis_period = {
+                "from": min(timestamps).strftime("%Y-%m-%d"),
+                "to": max(timestamps).strftime("%Y-%m-%d"),
+            }
+
+        return CrawlBudgetResult(
+            log_file=self.log_file,
+            analysis_period=analysis_period,
+            total_bot_requests=len(entries),
+            bots=bot_profiles,
+            waste=waste,
+            total_waste_pct=total_waste_pct,
+            orphan_pages=orphans,
+            recommendations=recommendations,
+            efficiency_score=efficiency_score,
+            timestamp=datetime.now().isoformat(),
+        )
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def main() -> None:
+    parser = argparse.ArgumentParser(
+        description="Analyze crawl budget efficiency and generate optimization recommendations.",
+    )
+    parser.add_argument(
+        "--log-file",
+        required=True,
+        help="Path to server access log file",
+    )
+    parser.add_argument(
+        "--sitemap",
+        default=None,
+        help="URL of XML sitemap for orphan page detection",
+    )
+    parser.add_argument(
+        "--url",
+        default=None,
+        help="Target website URL (used for URL normalization and Ahrefs)",
+    )
+    parser.add_argument(
+        "--scope",
+        choices=["all", "waste", "orphans", "bots"],
+        default="all",
+        help="Analysis scope (default: all)",
+    )
+    parser.add_argument(
+        "--ahrefs",
+        action="store_true",
+        help="Include Ahrefs page history comparison (requires MCP tool)",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        help="Output in JSON format",
+    )
+    parser.add_argument(
+        "--output",
+        default=None,
+        help="Write output to file instead of stdout",
+    )
+    args = parser.parse_args()
+
+    # Validate log file exists
+    if not Path(args.log_file).exists():
+        logger.error(f"Log file not found: {args.log_file}")
+        sys.exit(1)
+
+    analyzer = CrawlBudgetAnalyzer(
+        log_file=args.log_file,
+        sitemap_url=args.sitemap,
+        target_url=args.url,
+    )
+
+    result = analyzer.analyze(scope=args.scope)
+
+    if args.json:
+        output_data = result.to_dict()
+        output_str = json.dumps(output_data, indent=2, ensure_ascii=False)
+    else:
+        lines = _format_text_report(result)
+        output_str = "\n".join(lines)
+
+    if args.output:
+        Path(args.output).write_text(output_str, encoding="utf-8")
+        logger.info(f"Output written to {args.output}")
+    else:
+        print(output_str)
+
+
+def _format_text_report(result: CrawlBudgetResult) -> list[str]:
+    """Format the analysis result as a human-readable text report."""
+    lines = [
+        "=" * 70,
+        "Crawl Budget Analysis Report",
+        "=" * 70,
+        f"Log File: {result.log_file}",
+        f"Total Bot Requests: {result.total_bot_requests:,}",
+        f"Efficiency Score: {result.efficiency_score}/100",
+        f"Total Waste: {result.total_waste_pct:.1f}%",
+    ]
+    if result.analysis_period:
+        lines.append(
+            f"Period: {result.analysis_period.get('from', 'N/A')} ~ "
+            f"{result.analysis_period.get('to', 'N/A')}"
+        )
+    lines.append("")
+
+    # Bot profiles
+    if result.bots:
+        lines.append("-" * 60)
+        lines.append("Bot Profiles")
+        lines.append("-" * 60)
+        for name, profile in sorted(result.bots.items(), key=lambda x: -x[1].total_requests):
+            lines.append(f"\n  [{name.upper()}]")
+            lines.append(f"    Requests: {profile.total_requests:,}")
+            lines.append(f"    Unique URLs: {profile.unique_urls:,}")
+            lines.append(f"    Requests/Day: {profile.requests_per_day:,.1f}")
+            lines.append(f"    Days Active: {profile.days_active}")
+            lines.append(f"    Peak Hours: {profile.peak_hours}")
+            lines.append(f"    Status: {profile.status_breakdown}")
+        lines.append("")
+
+    # Waste breakdown
+    if result.waste:
+        lines.append("-" * 60)
+        lines.append("Crawl Waste Breakdown")
+        lines.append("-" * 60)
+        for w in result.waste:
+            if w.count > 0:
+                lines.append(f"\n  [{w.waste_type}]")
+                lines.append(f"    Count: {w.count:,} ({w.pct_of_total:.1f}%)")
+                lines.append(f"    Recommendation: {w.recommendation}")
+                if w.urls:
+                    lines.append(f"    Sample URLs:")
+                    for u in w.urls[:5]:
+                        lines.append(f"      - {u}")
+        lines.append("")
+
+    # Orphan pages
+    not_crawled = result.orphan_pages.get("in_sitemap_not_crawled", [])
+    not_in_sitemap = result.orphan_pages.get("crawled_not_in_sitemap", [])
+    if not_crawled or not_in_sitemap:
+        lines.append("-" * 60)
+        lines.append("Orphan Pages")
+        lines.append("-" * 60)
+        if not_crawled:
+            lines.append(f"\n  In Sitemap but Not Crawled: {len(not_crawled):,}")
+            for op in not_crawled[:10]:
+                lines.append(f"    - {op.url}")
+        if not_in_sitemap:
+            lines.append(f"\n  Crawled but Not in Sitemap: {len(not_in_sitemap):,}")
+            for op in not_in_sitemap[:10]:
+                lines.append(f"    - {op.url}")
+        lines.append("")
+
+    # Recommendations
+    if result.recommendations:
+        lines.append("-" * 60)
+        lines.append("Recommendations")
+        lines.append("-" * 60)
+        for i, rec in enumerate(result.recommendations, 1):
+            lines.append(f"\n  {i}. [{rec.priority.upper()}] {rec.category}")
+            lines.append(f"     Action: {rec.action}")
+            lines.append(f"     Impact: {rec.impact}")
+            lines.append(f"     Details: {rec.details}")
+
+    lines.append("")
+    lines.append(f"Generated: {result.timestamp}")
+    return lines
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/32-seo-crawl-budget/code/scripts/log_parser.py b/custom-skills/32-seo-crawl-budget/code/scripts/log_parser.py
new file mode 100644
index 0000000..36c230a
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/code/scripts/log_parser.py
@@ -0,0 +1,613 @@
+"""
+Log Parser - Server access log parser with bot identification
+=============================================================
+Purpose: Parse Apache/Nginx/CloudFront access logs, identify search engine
+         bots, extract crawl data, and generate per-bot statistics.
+Python: 3.10+
+"""
+
+import argparse
+import bz2
+import gzip
+import json
+import logging
+import re
+import sys
+from collections import Counter, defaultdict
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from pathlib import Path
+from typing import Generator, TextIO
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+)
+logger = logging.getLogger(__name__)
+
+
+# ---------------------------------------------------------------------------
+# Constants: bot user-agent patterns
+# ---------------------------------------------------------------------------
+
+BOT_PATTERNS: list[tuple[str, str, str]] = [
+    # (canonical name, regex pattern, category)
+    ("googlebot", r"Googlebot(?:-Image|-News|-Video)?/", "search_engine"),
+    ("googlebot-adsbot", r"AdsBot-Google", "search_engine"),
+    ("googlebot-mediapartners", r"Mediapartners-Google", "search_engine"),
+    ("yeti", r"Yeti/", "search_engine"),
+    ("bingbot", r"bingbot/", "search_engine"),
+    ("daumoa", r"Daumoa", "search_engine"),
+    ("applebot", r"Applebot/", "search_engine"),
+    ("duckduckbot", r"DuckDuckBot/", "search_engine"),
+    ("baiduspider", r"Baiduspider", "search_engine"),
+    ("yandexbot", r"YandexBot/", "search_engine"),
+    ("sogou", r"Sogou", "search_engine"),
+    ("seznambot", r"SeznamBot/", "search_engine"),
+    ("ahrefsbot", r"AhrefsBot/", "seo_tool"),
+    ("semrushbot", r"SemrushBot/", "seo_tool"),
+    ("mj12bot", r"MJ12bot/", "seo_tool"),
+    ("dotbot", r"DotBot/", "seo_tool"),
+    ("rogerbot", r"rogerbot/", "seo_tool"),
+    ("screaming frog", r"Screaming Frog SEO Spider", "seo_tool"),
+]
+
+COMPILED_BOT_PATTERNS: list[tuple[str, re.Pattern, str]] = [
+    (name, re.compile(pattern, re.IGNORECASE), category)
+    for name, pattern, category in BOT_PATTERNS
+]
+
+# ---------------------------------------------------------------------------
+# Regex patterns for each log format
+# ---------------------------------------------------------------------------
+
+NGINX_COMBINED_RE = re.compile(
+    r'(?P<ip>[\d.:a-fA-F]+)\s+-\s+(?P<user>\S+)\s+'
+    r'\[(?P<timestamp>[^\]]+)\]\s+'
+    r'"(?P<method>\S+)\s+(?P<url>\S+)\s+(?P<protocol>[^"]+)"\s+'
+    r'(?P<status>\d{3})\s+(?P<size>\d+|-)\s+'
+    r'"(?P<referer>[^"]*)"\s+'
+    r'"(?P<user_agent>[^"]*)"'
+)
+
+APACHE_COMBINED_RE = re.compile(
+    r'(?P<ip>[\d.:a-fA-F]+)\s+\S+\s+(?P<user>\S+)\s+'
+    r'\[(?P<timestamp>[^\]]+)\]\s+'
+    r'"(?P<method>\S+)\s+(?P<url>\S+)\s+(?P<protocol>[^"]+)"\s+'
+    r'(?P<status>\d{3})\s+(?P<size>\d+|-)\s+'
+    r'"(?P<referer>[^"]*)"\s+'
+    r'"(?P<user_agent>[^"]*)"'
+)
+
+CLOUDFRONT_FIELDS = [
+    "date", "time", "x_edge_location", "sc_bytes", "c_ip",
+    "cs_method", "cs_host", "cs_uri_stem", "sc_status",
+    "cs_referer", "cs_user_agent", "cs_uri_query",
+    "cs_cookie", "x_edge_result_type", "x_edge_request_id",
+    "x_host_header", "cs_protocol", "cs_bytes",
+    "time_taken", "x_forwarded_for", "ssl_protocol",
+    "ssl_cipher", "x_edge_response_result_type", "cs_protocol_version",
+]
+
+# Timestamp formats
+NGINX_TS_FORMAT = "%d/%b/%Y:%H:%M:%S %z"
+APACHE_TS_FORMAT = "%d/%b/%Y:%H:%M:%S %z"
+
+
+# ---------------------------------------------------------------------------
+# Data classes
+# ---------------------------------------------------------------------------
+
+@dataclass
+class LogEntry:
+    """A single parsed log entry."""
+    timestamp: datetime | None
+    ip: str
+    method: str
+    url: str
+    status_code: int
+    response_size: int
+    user_agent: str
+    referer: str
+
+    def to_dict(self) -> dict:
+        d = asdict(self)
+        if self.timestamp:
+            d["timestamp"] = self.timestamp.isoformat()
+        return d
+
+
+@dataclass
+class BotIdentification:
+    """Bot identification result."""
+    name: str
+    user_agent_pattern: str
+    category: str
+
+
+@dataclass
+class BotStats:
+    """Aggregated statistics for a single bot."""
+    name: str
+    total_requests: int = 0
+    unique_urls: int = 0
+    status_distribution: dict[str, int] = field(default_factory=dict)
+    top_urls: list[tuple[str, int]] = field(default_factory=list)
+    hourly_distribution: dict[int, int] = field(default_factory=dict)
+    daily_distribution: dict[str, int] = field(default_factory=dict)
+    avg_response_size: float = 0.0
+
+    def to_dict(self) -> dict:
+        return {
+            "name": self.name,
+            "total_requests": self.total_requests,
+            "unique_urls": self.unique_urls,
+            "status_distribution": self.status_distribution,
+            "top_urls": [{"url": u, "count": c} for u, c in self.top_urls],
+            "hourly_distribution": self.hourly_distribution,
+            "daily_distribution": self.daily_distribution,
+            "avg_response_size": round(self.avg_response_size, 1),
+        }
+
+
+@dataclass
+class LogParseResult:
+    """Complete log parsing result."""
+    log_file: str
+    format_detected: str
+    total_lines: int
+    parsed_lines: int
+    bot_entries: int
+    date_range: dict[str, str]
+    bots: dict[str, BotStats]
+    errors: int
+
+    def to_dict(self) -> dict:
+        return {
+            "log_file": self.log_file,
+            "format_detected": self.format_detected,
+            "total_lines": self.total_lines,
+            "parsed_lines": self.parsed_lines,
+            "bot_entries": self.bot_entries,
+            "date_range": self.date_range,
+            "bots": {name: stats.to_dict() for name, stats in self.bots.items()},
+            "errors": self.errors,
+        }
+
+
+# ---------------------------------------------------------------------------
+# LogParser class
+# ---------------------------------------------------------------------------
+
+class LogParser:
+    """Parse server access logs and identify search engine bot traffic."""
+
+    def __init__(
+        self,
+        log_file: str,
+        fmt: str = "auto",
+        streaming: bool = False,
+    ):
+        self.log_file = log_file
+        self.fmt = fmt
+        self.streaming = streaming
+        self._detected_format: str | None = None
+        self._parse_errors = 0
+
+    # -- format detection -----------------------------------------------------
+
+    def _detect_format(self, line: str) -> str:
+        """Auto-detect log format from a sample line."""
+        if line.startswith("#"):
+            return "cloudfront"
+        if NGINX_COMBINED_RE.match(line):
+            return "nginx"
+        if APACHE_COMBINED_RE.match(line):
+            return "apache"
+        # Fallback: try tab-separated (CloudFront without header)
+        if "\t" in line and line.count("\t") >= 10:
+            return "cloudfront"
+        return "nginx"
+
+    # -- line parsers ---------------------------------------------------------
+
+    def _parse_nginx_combined(self, line: str) -> LogEntry | None:
+        """Parse a single Nginx combined format log line."""
+        m = NGINX_COMBINED_RE.match(line)
+        if not m:
+            return None
+        ts = None
+        try:
+            ts = datetime.strptime(m.group("timestamp"), NGINX_TS_FORMAT)
+        except (ValueError, TypeError):
+            pass
+        size_raw = m.group("size")
+        size = int(size_raw) if size_raw != "-" else 0
+        return LogEntry(
+            timestamp=ts,
+            ip=m.group("ip"),
+            method=m.group("method"),
+            url=m.group("url"),
+            status_code=int(m.group("status")),
+            response_size=size,
+            user_agent=m.group("user_agent"),
+            referer=m.group("referer"),
+        )
+
+    def _parse_apache_combined(self, line: str) -> LogEntry | None:
+        """Parse a single Apache combined format log line."""
+        m = APACHE_COMBINED_RE.match(line)
+        if not m:
+            return None
+        ts = None
+        try:
+            ts = datetime.strptime(m.group("timestamp"), APACHE_TS_FORMAT)
+        except (ValueError, TypeError):
+            pass
+        size_raw = m.group("size")
+        size = int(size_raw) if size_raw != "-" else 0
+        return LogEntry(
+            timestamp=ts,
+            ip=m.group("ip"),
+            method=m.group("method"),
+            url=m.group("url"),
+            status_code=int(m.group("status")),
+            response_size=size,
+            user_agent=m.group("user_agent"),
+            referer=m.group("referer"),
+        )
+
+    def _parse_cloudfront(self, line: str) -> LogEntry | None:
+        """Parse a CloudFront tab-separated log line."""
+        if line.startswith("#"):
+            return None
+        parts = line.strip().split("\t")
+        if len(parts) < 13:
+            return None
+        ts = None
+        try:
+            ts = datetime.strptime(f"{parts[0]} {parts[1]}", "%Y-%m-%d %H:%M:%S")
+        except (ValueError, IndexError):
+            pass
+        try:
+            status = int(parts[8])
+        except (ValueError, IndexError):
+            status = 0
+        try:
+            size = int(parts[3])
+        except (ValueError, IndexError):
+            size = 0
+        url = parts[7] if len(parts) > 7 else ""
+        query = parts[11] if len(parts) > 11 else ""
+        if query and query != "-":
+            url = f"{url}?{query}"
+        ua = parts[10] if len(parts) > 10 else ""
+        ua = ua.replace("%20", " ").replace("%2520", " ")
+        referer = parts[9] if len(parts) > 9 else ""
+        return LogEntry(
+            timestamp=ts,
+            ip=parts[4] if len(parts) > 4 else "",
+            method=parts[5] if len(parts) > 5 else "",
+            url=url,
+            status_code=status,
+            response_size=size,
+            user_agent=ua,
+            referer=referer,
+        )
+
+    def _parse_line(self, line: str, fmt: str) -> LogEntry | None:
+        """Route to the correct parser based on format."""
+        parsers = {
+            "nginx": self._parse_nginx_combined,
+            "apache": self._parse_apache_combined,
+            "cloudfront": self._parse_cloudfront,
+        }
+        parser = parsers.get(fmt, self._parse_nginx_combined)
+        return parser(line)
+
+    # -- bot identification ---------------------------------------------------
+
+    @staticmethod
+    def identify_bot(user_agent: str) -> BotIdentification | None:
+        """Match user-agent against known bot patterns."""
+        if not user_agent or user_agent == "-":
+            return None
+        for name, pattern, category in COMPILED_BOT_PATTERNS:
+            if pattern.search(user_agent):
+                return BotIdentification(
+                    name=name,
+                    user_agent_pattern=pattern.pattern,
+                    category=category,
+                )
+        # Heuristic: generic bot detection via common keywords
+        ua_lower = user_agent.lower()
+        bot_keywords = ["bot", "spider", "crawler", "scraper", "fetch"]
+        for kw in bot_keywords:
+            if kw in ua_lower:
+                return BotIdentification(
+                    name="other",
+                    user_agent_pattern=kw,
+                    category="other",
+                )
+        return None
+
+    # -- file handling --------------------------------------------------------
+
+    @staticmethod
+    def _open_file(path: str) -> TextIO:
+        """Open plain text, .gz, or .bz2 log files."""
+        p = Path(path)
+        if p.suffix == ".gz":
+            return gzip.open(path, "rt", encoding="utf-8", errors="replace")
+        if p.suffix == ".bz2":
+            return bz2.open(path, "rt", encoding="utf-8", errors="replace")
+        return open(path, "r", encoding="utf-8", errors="replace")
+
+    # -- streaming parser -----------------------------------------------------
+
+    def parse_streaming(
+        self,
+        filter_bot: str | None = None,
+    ) -> Generator[tuple[LogEntry, BotIdentification], None, None]:
+        """Generator-based streaming parser for large files."""
+        fmt = self.fmt
+        first_line_checked = False
+
+        fh = self._open_file(self.log_file)
+        try:
+            for line in fh:
+                line = line.strip()
+                if not line:
+                    continue
+                if not first_line_checked and fmt == "auto":
+                    fmt = self._detect_format(line)
+                    self._detected_format = fmt
+                    first_line_checked = True
+                entry = self._parse_line(line, fmt)
+                if entry is None:
+                    self._parse_errors += 1
+                    continue
+                bot = self.identify_bot(entry.user_agent)
+                if bot is None:
+                    continue
+                if filter_bot and bot.name != filter_bot.lower():
+                    continue
+                yield entry, bot
+        finally:
+            fh.close()
+
+    # -- full parse -----------------------------------------------------------
+
+    def parse(
+        self,
+        filter_bot: str | None = None,
+        date_from: datetime | None = None,
+        date_to: datetime | None = None,
+    ) -> list[tuple[LogEntry, BotIdentification]]:
+        """Full parse with optional date and bot filters."""
+        results: list[tuple[LogEntry, BotIdentification]] = []
+        for entry, bot in self.parse_streaming(filter_bot):
+            if date_from and entry.timestamp and entry.timestamp < date_from:
+                continue
+            if date_to and entry.timestamp and entry.timestamp > date_to:
+                continue
+            results.append((entry, bot))
+        return results
+
+    # -- statistics -----------------------------------------------------------
+
+    @staticmethod
+    def get_bot_stats(
+        entries: list[tuple[LogEntry, BotIdentification]],
+    ) -> dict[str, BotStats]:
+        """Aggregate per-bot statistics from parsed entries."""
+        bot_data: dict[str, dict] = defaultdict(lambda: {
+            "urls": Counter(),
+            "statuses": Counter(),
+            "hours": Counter(),
+            "days": Counter(),
+            "sizes": [],
+            "count": 0,
+        })
+
+        for entry, bot in entries:
+            bd = bot_data[bot.name]
+            bd["count"] += 1
+            bd["urls"][entry.url] += 1
+            bd["statuses"][str(entry.status_code)] += 1
+            bd["sizes"].append(entry.response_size)
+            if entry.timestamp:
+                bd["hours"][entry.timestamp.hour] += 1
+                day_key = entry.timestamp.strftime("%Y-%m-%d")
+                bd["days"][day_key] += 1
+
+        stats: dict[str, BotStats] = {}
+        for name, bd in bot_data.items():
+            avg_size = sum(bd["sizes"]) / len(bd["sizes"]) if bd["sizes"] else 0.0
+            top_20 = bd["urls"].most_common(20)
+            stats[name] = BotStats(
+                name=name,
+                total_requests=bd["count"],
+                unique_urls=len(bd["urls"]),
+                status_distribution=dict(bd["statuses"]),
+                top_urls=top_20,
+                hourly_distribution=dict(sorted(bd["hours"].items())),
+                daily_distribution=dict(sorted(bd["days"].items())),
+                avg_response_size=avg_size,
+            )
+        return stats
+
+    # -- orchestrator ---------------------------------------------------------
+
+    def parse_and_analyze(
+        self,
+        filter_bot: str | None = None,
+        date_from: datetime | None = None,
+        date_to: datetime | None = None,
+    ) -> LogParseResult:
+        """Orchestrate parsing and statistics generation."""
+        entries = self.parse(filter_bot, date_from, date_to)
+        bot_stats = self.get_bot_stats(entries)
+
+        # Determine date range
+        timestamps = [e.timestamp for e, _ in entries if e.timestamp]
+        date_range = {}
+        if timestamps:
+            date_range = {
+                "from": min(timestamps).isoformat(),
+                "to": max(timestamps).isoformat(),
+            }
+
+        # Count total lines for context
+        total_lines = 0
+        fh = self._open_file(self.log_file)
+        try:
+            for _ in fh:
+                total_lines += 1
+        finally:
+            fh.close()
+
+        return LogParseResult(
+            log_file=self.log_file,
+            format_detected=self._detected_format or self.fmt,
+            total_lines=total_lines,
+            parsed_lines=total_lines - self._parse_errors,
+            bot_entries=len(entries),
+            date_range=date_range,
+            bots=bot_stats,
+            errors=self._parse_errors,
+        )
+
+
+# ---------------------------------------------------------------------------
+# CLI
+# ---------------------------------------------------------------------------
+
+def _parse_date(val: str) -> datetime:
+    """Parse a date string in YYYY-MM-DD format."""
+    return datetime.strptime(val, "%Y-%m-%d")
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(
+        description="Parse server access logs and identify search engine bot traffic.",
+    )
+    parser.add_argument(
+        "--log-file",
+        required=True,
+        help="Path to access log file (plain, .gz, .bz2)",
+    )
+    parser.add_argument(
+        "--format",
+        dest="fmt",
+        choices=["auto", "nginx", "apache", "cloudfront"],
+        default="auto",
+        help="Log format (default: auto-detect)",
+    )
+    parser.add_argument(
+        "--bot",
+        default=None,
+        help="Filter results to a specific bot (e.g., googlebot, yeti, bingbot, daumoa)",
+    )
+    parser.add_argument(
+        "--streaming",
+        action="store_true",
+        help="Use streaming parser for large files (prints entries incrementally)",
+    )
+    parser.add_argument(
+        "--date-from",
+        default=None,
+        help="Filter entries from date (YYYY-MM-DD)",
+    )
+    parser.add_argument(
+        "--date-to",
+        default=None,
+        help="Filter entries to date (YYYY-MM-DD)",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        help="Output in JSON format",
+    )
+    parser.add_argument(
+        "--output",
+        default=None,
+        help="Write output to file instead of stdout",
+    )
+    args = parser.parse_args()
+
+    # Validate file exists
+    if not Path(args.log_file).exists():
+        logger.error(f"Log file not found: {args.log_file}")
+        sys.exit(1)
+
+    date_from = _parse_date(args.date_from) if args.date_from else None
+    date_to = _parse_date(args.date_to) if args.date_to else None
+
+    lp = LogParser(log_file=args.log_file, fmt=args.fmt, streaming=args.streaming)
+
+    if args.streaming and not args.json:
+        # Streaming mode: print entries as they are parsed
+        count = 0
+        for entry, bot in lp.parse_streaming(args.bot):
+            if date_from and entry.timestamp and entry.timestamp < date_from:
+                continue
+            if date_to and entry.timestamp and entry.timestamp > date_to:
+                continue
+            ts_str = entry.timestamp.isoformat() if entry.timestamp else "N/A"
+            print(
+                f"[{bot.name}] {ts_str} {entry.status_code} "
+                f"{entry.method} {entry.url} ({entry.response_size}B)"
+            )
+            count += 1
+        print(f"\n--- Total bot requests: {count} ---")
+        return
+
+    # Full analysis mode
+    result = lp.parse_and_analyze(
+        filter_bot=args.bot,
+        date_from=date_from,
+        date_to=date_to,
+    )
+
+    if args.json:
+        output_data = result.to_dict()
+        output_str = json.dumps(output_data, indent=2, ensure_ascii=False)
+    else:
+        lines = [
+            f"Log File: {result.log_file}",
+            f"Format: {result.format_detected}",
+            f"Total Lines: {result.total_lines:,}",
+            f"Parsed Lines: {result.parsed_lines:,}",
+            f"Bot Entries: {result.bot_entries:,}",
+            f"Parse Errors: {result.errors:,}",
+        ]
+        if result.date_range:
+            lines.append(f"Date Range: {result.date_range.get('from', 'N/A')} to {result.date_range.get('to', 'N/A')}")
+        lines.append("")
+        lines.append("=" * 60)
+        lines.append("Bot Statistics")
+        lines.append("=" * 60)
+        for name, stats in sorted(result.bots.items(), key=lambda x: -x[1].total_requests):
+            lines.append(f"\n--- {name.upper()} ---")
+            lines.append(f"  Requests: {stats.total_requests:,}")
+            lines.append(f"  Unique URLs: {stats.unique_urls:,}")
+            lines.append(f"  Avg Response Size: {stats.avg_response_size:,.0f} bytes")
+            lines.append(f"  Status Distribution: {stats.status_distribution}")
+            lines.append(f"  Top 10 URLs:")
+            for url, cnt in stats.top_urls[:10]:
+                lines.append(f"    {cnt:>6,} | {url}")
+            if stats.hourly_distribution:
+                peak_hour = max(stats.hourly_distribution, key=stats.hourly_distribution.get)
+                lines.append(f"  Peak Hour: {peak_hour}:00 ({stats.hourly_distribution[peak_hour]:,} reqs)")
+        output_str = "\n".join(lines)
+
+    if args.output:
+        Path(args.output).write_text(output_str, encoding="utf-8")
+        logger.info(f"Output written to {args.output}")
+    else:
+        print(output_str)
+
+
+if __name__ == "__main__":
+    main()
diff --git a/custom-skills/32-seo-crawl-budget/code/scripts/requirements.txt b/custom-skills/32-seo-crawl-budget/code/scripts/requirements.txt
new file mode 100644
index 0000000..ef6ec2e
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/code/scripts/requirements.txt
@@ -0,0 +1,10 @@
+# 32-seo-crawl-budget dependencies
+requests>=2.31.0
+aiohttp>=3.9.0
+pandas>=2.1.0
+beautifulsoup4>=4.12.0
+lxml>=5.1.0
+tenacity>=8.2.0
+tqdm>=4.66.0
+python-dotenv>=1.0.0
+rich>=13.7.0
diff --git a/custom-skills/32-seo-crawl-budget/desktop/SKILL.md b/custom-skills/32-seo-crawl-budget/desktop/SKILL.md
new file mode 100644
index 0000000..e6cee04
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/desktop/SKILL.md
@@ -0,0 +1,39 @@
+---
+name: seo-crawl-budget
+description: |
+  Crawl budget optimization and log analysis. Triggers: crawl budget, log analysis, bot crawling, Googlebot, crawl waste, orphan pages, crawl efficiency.
+---
+
+# Crawl Budget Optimizer
+
+Analyze server access logs to identify crawl budget waste and generate optimization recommendations for search engine bots.
+
+## Capabilities
+
+1. **Log Analysis**: Parse Nginx/Apache/CloudFront access logs to extract bot crawl data
+2. **Bot Profiling**: Per-bot behavior analysis (Googlebot, Yeti, Bingbot, Daumoa)
+3. **Waste Detection**: Parameter URLs, redirect chains, soft 404s, duplicate URL variants
+4. **Orphan Pages**: Pages in sitemap but uncrawled, and crawled pages not in sitemap
+5. **Recommendations**: Prioritized action items for crawl budget optimization
+
+## Workflow
+
+1. Parse server access log with `log_parser.py`
+2. Run crawl budget analysis with `crawl_budget_analyzer.py`
+3. Compare with sitemap URLs for orphan page detection
+4. Optionally compare with Ahrefs page history data
+5. Generate Korean-language report with recommendations
+6. Save to Notion SEO Audit Log database
+
+## Tools Used
+
+- **Ahrefs**: `site-explorer-pages-history` for indexed page comparison
+- **Notion**: Save audit report to database `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+- **WebSearch**: Current best practices and bot documentation
+
+## Output
+
+All reports are saved to the OurDigital SEO Audit Log with:
+- Category: Crawl Budget
+- Audit ID format: CRAWL-YYYYMMDD-NNN
+- Content in Korean with technical English terms preserved
diff --git a/custom-skills/32-seo-crawl-budget/desktop/skill.yaml b/custom-skills/32-seo-crawl-budget/desktop/skill.yaml
new file mode 100644
index 0000000..fb443ca
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/desktop/skill.yaml
@@ -0,0 +1,8 @@
+name: seo-crawl-budget
+description: |
+  Crawl budget optimization and log analysis. Triggers: crawl budget, log analysis, bot crawling, Googlebot, crawl waste, orphan pages, crawl efficiency.
+allowed-tools:
+  - mcp__ahrefs__*
+  - mcp__notion__*
+  - WebSearch
+  - WebFetch
diff --git a/custom-skills/32-seo-crawl-budget/desktop/tools/ahrefs.md b/custom-skills/32-seo-crawl-budget/desktop/tools/ahrefs.md
new file mode 100644
index 0000000..dcb8687
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/desktop/tools/ahrefs.md
@@ -0,0 +1,17 @@
+# Ahrefs MCP Tools
+
+## site-explorer-pages-history
+
+Get historical page data for a domain to compare indexed pages with crawled pages.
+
+```
+mcp__ahrefs__site-explorer-pages-history
+```
+
+**Parameters**:
+- `target` (string, required): Domain or URL to analyze
+- `date_from` (string): Start date (YYYY-MM-DD)
+- `date_to` (string): End date (YYYY-MM-DD)
+- `mode` (string): "domain", "prefix", "exact", "subdomains"
+
+**Use case**: Compare Ahrefs indexed page counts with server log crawl data to identify indexing gaps and crawl budget inefficiencies.
diff --git a/custom-skills/32-seo-crawl-budget/desktop/tools/notion.md b/custom-skills/32-seo-crawl-budget/desktop/tools/notion.md
new file mode 100644
index 0000000..afc4783
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/desktop/tools/notion.md
@@ -0,0 +1,21 @@
+# Notion MCP Tools
+
+## notion-create-pages
+
+Create a new page in the OurDigital SEO Audit Log database.
+
+```
+mcp__notion__notion-create-pages
+```
+
+**Database ID**: `2c8581e5-8a1e-8035-880b-e38cefc2f3ef`
+
+**Required Properties**:
+- `Issue` (Title): Report title in Korean with date
+- `Site` (URL): Audited website URL
+- `Category` (Select): "Crawl Budget"
+- `Priority` (Select): Based on waste percentage (Critical >20%, High >10%, Medium >5%, Low <5%)
+- `Found Date` (Date): Audit date (YYYY-MM-DD)
+- `Audit ID` (Rich Text): Format CRAWL-YYYYMMDD-NNN
+
+**Content**: Full crawl budget report in Korean with technical English terms preserved.
diff --git a/custom-skills/32-seo-crawl-budget/desktop/tools/websearch.md b/custom-skills/32-seo-crawl-budget/desktop/tools/websearch.md
new file mode 100644
index 0000000..d17c9cf
--- /dev/null
+++ b/custom-skills/32-seo-crawl-budget/desktop/tools/websearch.md
@@ -0,0 +1,18 @@
+# WebSearch Tool
+
+## Purpose
+
+Search the web for current crawl budget best practices, search engine bot documentation, and robots.txt guidelines.
+
+## Usage
+
+```
+WebSearch(query="Googlebot crawl budget optimization 2025")
+```
+
+**Common queries**:
+- Search engine bot crawl rate documentation
+- robots.txt best practices for crawl budget
+- URL parameter handling for search engines
+- Crawl budget optimization techniques
+- Search engine bot user-agent strings