our-claude-skills/custom-skills/32-notion-writer/code/CLAUDE.md

# Notion Writer - Claude Code Skill

> **Purpose**: Push markdown content to Notion pages or databases
> **Platform**: Claude Code (CLI)
> **Input**: Markdown files, Notion URLs
> **Output**: Content written to Notion

---

## Capabilities

| Feature | Input | Output |
|---------|-------|--------|
| Page Content Append | Markdown + Page URL | Appended blocks |
| Page Content Replace | Markdown + Page URL | Replaced content |
| Database Row Create | Markdown + DB URL + Title | New database row |
| Connection Test | API token | Connection status |
| Page/DB Info | URL | Metadata |

---

## Setup

### 1. Create Notion Integration

1. Go to [Notion Integrations](https://www.notion.so/my-integrations)
2. Click "New integration"
3. Name it (e.g., "Claude Writer")
4. Select workspace
5. Copy the "Internal Integration Token"

### 2. Share Pages/Databases with Integration

**Important**: You must share each page or database with your integration:
1. Open the Notion page/database
2. Click "..." menu → "Connections"
3. Add your integration

### 3. Configure Environment

```bash
cd ~/Project/our-claude-skills/custom-skills/32-notion-writer/code/scripts

# Create .env from example
cp .env.example .env

# Edit and add your token
nano .env
```

`.env` file:
```
NOTION_API_KEY=secret_your_integration_token
```

### 4. Activate Environment

```bash
source venv/bin/activate
```

---

## Usage

### Test Connection

```bash
python notion_writer.py --test
```

### Get Page/Database Info

```bash
# Page info
python notion_writer.py --page "https://notion.so/My-Page-abc123" --info

# Database info
python notion_writer.py --database "https://notion.so/abc123" --info
```

### Write to Page

```bash
# Append content to page
python notion_writer.py --page PAGE_URL --file content.md

# Replace page content
python notion_writer.py --page PAGE_URL --file content.md --replace

# From stdin
cat report.md | python notion_writer.py --page PAGE_URL --stdin
```

### Create Database Row

```bash
# Create row with title and content
python notion_writer.py --database DB_URL --title "New Entry" --file content.md

# Title only
python notion_writer.py --database DB_URL --title "Empty Entry"

# Set arbitrary properties (JSON literal)
python notion_writer.py \
  --database DB_URL \
  --title "Research Notes" \
  --properties '{"Status": "In progress", "Topic": ["AI", "MCP"], "Account Code": "OurDigital"}' \
  --file notes.md

# Set properties from a JSON file
python notion_writer.py \
  --database DB_URL \
  --title "Research Notes" \
  --properties ./props.json \
  --file notes.md
```

`--database` accepts a database URL, a database_id, or a data_source_id. It is resolved to a data_source_id automatically.

### Upsert (avoid duplicates)

```bash
# If a row with Name="My Title" already exists, update it; otherwise create.
python notion_writer.py \
  --database DB_URL \
  --title "My Title" \
  --upsert-by "Name" \
  --properties '{"Status": "Done"}' \
  --file content.md
```

When upserting, `--file` (or `--stdin`) replaces the page's body blocks; `--properties` updates the named columns. `--upsert-by` supports title, rich_text, select, status, number, url, email, and phone_number properties.

### Two-tool split: when to use what

| Goal | Tool | Why |
|------|------|-----|
| Push markdown content into a page or new DB row | **`notion_writer.py`** | Converts markdown → Notion blocks (tables, headings, code, lists, rich-text, anchors) |
| Create a DB row with both content and metadata in one shot | **`notion_writer.py`** with `--properties` | Single API call, full schema validation |
| Update properties on many existing pages (bulk column edit) | **MCP `notion-update-page`** | Native schema awareness, no markdown work needed |
| Search / move pages with metadata transition | 31-notion-organizer (Phase 3 work) | Cross-database operations |

---

## Markdown Support

### Supported Elements

| Markdown | Notion Block |
|----------|--------------|
| `# Heading` | Heading 1 |
| `## Heading` | Heading 2 |
| `### Heading` | Heading 3 |
| `- item` | Bulleted list |
| `1. item` | Numbered list |
| `- [ ] task` | To-do (unchecked) |
| `- [x] task` | To-do (checked) |
| `> quote` | Quote |
| `> [!NOTE]` / `[!TIP]` / `[!IMPORTANT]` / `[!WARNING]` / `[!CAUTION]` | Callout (with icon + colored background) |
| `<details><summary>X</summary> ... </details>` | Toggle (multi-line form, recursive children) |
| `::: columns / ::: column / :::` | Column list (Pandoc fenced div, ≥2 columns required) |
| `` ```code``` `` | Code block |
| `---` | Divider |
| Tables (`\| col \|`) | Table |
| Paragraphs | Paragraph |

### Code Block Languages

Specify language after opening backticks:
```markdown
```python
print("Hello")
```
```

### Inline rich-text

| Markdown | Result |
|----------|--------|
| `**bold**` or `__bold__` | bold |
| `*italic*` or `_italic_` | italic |
| `` `code` `` | inline code |
| `~~strike~~` | strikethrough |
| `[text](https://...)` | link (absolute URLs only) |
| `[text](#anchor)` | bold (Notion rejects fragment URLs) |
| `[text](relative/path.md)` | plain text (Notion rejects relative URLs) |
| `@[Title](page-id-or-notion-url)` | page mention (resolves via Notion URL or raw 32-hex ID) |

Notion's URL validator requires absolute URLs for link annotations. The parser converts TOC-style anchor links to bold to preserve navigation intent and silently strips relative paths.

---

## Examples

### Push SEO Audit Report

```bash
python notion_writer.py \
  --page "https://notion.so/SEO-Reports-abc123" \
  --file ~/reports/seo_audit_2025.md
```

### Create Meeting Notes Entry

```bash
python notion_writer.py \
  --database "https://notion.so/Meeting-Notes-abc123" \
  --title "Weekly Standup - Dec 26" \
  --file meeting_notes.md
```

### Callouts (GitHub alerts)

````markdown
> [!NOTE]
> Just FYI: this method is idempotent.

> [!WARNING]
> Don't run this in production without a backup.
````

Renders as Notion callout blocks with corresponding emoji icon and colored background.

### Toggles (HTML5 `<details>`)

````markdown
<details>
<summary>Click to expand: full debug log</summary>

```bash
$ python notion_writer.py --test
✅ Connected
```

Lists, code blocks, and even nested `<details>` work inside.
</details>
````

Multi-line form required: `<details>` and `</details>` must be on their own lines.

### Columns (Pandoc fenced div)

````markdown
::: columns
::: column
**Column 1**

- item a
- item b
:::
::: column
**Column 2**

```python
print("hello")
```
:::
:::
````

≥2 columns required by Notion. Single-column blocks degrade to plain paragraphs.

### Page mentions

```markdown
See @[Architecture Decision Record](https://notion.so/ADR-abcdef0123456789abcdef0123456789) for context.
```

Both Notion URLs and raw 32-hex IDs work. Invalid targets fall back to plain text `@Title`.

### Pipe from Another Tool

```bash
# Pipe YouTube video info to Notion
python jamie_video_info.py VIDEO_ID --json | \
  python notion_writer.py --page PAGE_URL --stdin
```

### Test against the "Working with AI" database

The user's primary export target is the AI research database (data_source_id `f8f19ede-32bd-43ac-9f60-0651f6f40afe`). End-to-end check:

```bash
# 1) Sanity: connection works
python notion_writer.py --test

# 2) Inspect the schema (confirms data_source resolution)
python notion_writer.py \
  --database f8f19ede-32bd-43ac-9f60-0651f6f40afe \
  --info

# 3) Upsert a test row with full metadata
python notion_writer.py \
  --database f8f19ede-32bd-43ac-9f60-0651f6f40afe \
  --title "[smoke test] notion_writer v3 migration" \
  --upsert-by "Name" \
  --properties '{
    "Status": "Done",
    "AI used": ["Claude Code"],
    "Topic": ["AI"],
    "Category": ["Research"],
    "Account Code": "OurDigital",
    "Type": "Memo",
    "AI summary": "Verifying the data_sources API migration end-to-end."
  }' \
  --stdin <<< "## Smoke test

Migration verified — properties + content land in one shot."
```

Re-running step 3 should update the existing row instead of creating a duplicate (upsert).

---

## API Limits

| Limit | Value |
|-------|-------|
| Blocks per request | 100 |
| Text content per block | 2,000 chars |
| Requests per second | ~3 |

The script automatically batches large content.

---

## Troubleshooting

### "Notion object not found"
- Ensure the page or database is shared with your integration (Notion → "..." → Connections → add integration)
- Check URL/ID is correct
- For databases, the integration must have access to *each* data source under it

### "Unauthorized"
- Verify `NOTION_API_KEY` in `.env`
- Token should start with `secret_` (internal integration token)

### "Database has multiple data sources"
- Pass a specific `data_source_id` instead of the database URL/id
- List data sources via the Notion API or open the data source individually in Notion to grab its ID

### "Property 'X' has unsupported type"
- The script handles common property types but skips computed/read-only ones (formula, rollup, created_by, last_edited_by, last_edited_time, unique_id)
- Drop the unsupported key from `--properties` JSON

### "Rate limited"
- Wait and retry
- Script handles batching but rapid calls may hit limits

---

## File Structure

```
32-notion-writer/
├── code/
│   ├── CLAUDE.md              # This skill document
│   ├── scripts/
│   │   ├── notion_writer.py   # Main script
│   │   ├── _notion_compat.py  # API compat helper (data_sources resolution)
│   │   ├── test_parser.py     # Markdown parser tests
│   │   ├── venv/              # Python environment
│   │   ├── .env               # API token (not committed)
│   │   └── .env.example       # Template
│   ├── output/                # For generated content
│   └── references/
└── desktop/                   # Claude Desktop version (future)
```

---

## Quick Reference

```bash
# Navigate
cd ~/Project/our-claude-skills/custom-skills/32-notion-writer/code/scripts
source venv/bin/activate

# Test
python notion_writer.py --test

# Write to page
python notion_writer.py -p PAGE_URL -f content.md

# Replace content
python notion_writer.py -p PAGE_URL -f content.md -r

# Create DB row
python notion_writer.py -d DB_URL -t "Title" -f content.md

# Create DB row with full metadata
python notion_writer.py -d DB_URL -t "Title" --properties '{"Status": "Done"}' -f content.md

# Upsert (avoid duplicates)
python notion_writer.py -d DB_URL -t "Title" --upsert-by "Name" -f content.md
```

---

*Version 1.2.0 | Claude Code | 2026-04-27*

Changelog:
- 1.2.0 — Extended block coverage: GitHub-alert callouts, HTML5 `<details>` toggles, Pandoc `::: columns` fenced div, inline `@[Title](id-or-url)` page mentions. Parser made reentrant to support full recursion inside container blocks.
- 1.1.0 — Migrated to Notion API 2025-09-03 (multi-source databases). Added `--properties` JSON flag, `--upsert-by` for idempotency, anchor-link parser fix, friendlier API error messages.
- 1.0.0 — Initial release with markdown→Notion block conversion.