feat(reference-curator): Add pipeline orchestrator and refactor skill format

Pipeline Orchestrator:
- Add 07-pipeline-orchestrator skill with code/CLAUDE.md and desktop/SKILL.md
- Add /reference-curator-pipeline slash command for full workflow automation
- Add pipeline_runs and pipeline_iteration_tracker tables to schema.sql
- Add v_pipeline_status and v_pipeline_iterations views
- Add pipeline_config.yaml configuration template
- Update AGENTS.md with Reference Curator Skills section
- Update claude-project files with pipeline documentation

Skill Format Refactoring:
- Extract YAML frontmatter from SKILL.md files to separate skill.yaml
- Add tools/ directories with MCP tool documentation
- Update SKILL-FORMAT-REQUIREMENTS.md with new structure
- Add migrate-skill-structure.py script for format conversion

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
2026-01-29 01:01:02 +07:00
parent 243b9d851c
commit d1cd1298a8
91 changed files with 2475 additions and 281 deletions

View File

@@ -73,7 +73,24 @@ This is a Claude Skills collection repository containing:
|---|-------|---------|---------|
| 90 | reference-curator | Modular skill suite for reference documentation | See sub-skills below |
**Sub-skills (installed via `./install.sh`):**
**Pipeline Orchestrator (Recommended):**
| Command | Purpose |
|---------|---------|
| `/reference-curator-pipeline` | Full pipeline orchestration with QA loop handling |
```
# Full pipeline from topic
/reference-curator-pipeline "Claude Code best practices" --max-sources 5
# From URLs (skip discovery)
/reference-curator-pipeline https://docs.anthropic.com/en/docs/prompt-caching
# With auto-approve and fine-tuning output
/reference-curator-pipeline "MCP servers" --auto-approve --export-format fine_tuning
```
**Individual Sub-skills (installed via `./install.sh`):**
| Skill | Command | Purpose |
|-------|---------|---------|
@@ -83,6 +100,7 @@ This is a Claude Skills collection repository containing:
| content-distiller | `/content-distiller` | Summarization & key concept extraction |
| quality-reviewer | `/quality-reviewer` | QA loop with approve/refactor/research routing |
| markdown-exporter | `/markdown-exporter` | Export to markdown or JSONL for fine-tuning |
| pipeline-orchestrator | `/reference-curator-pipeline` | Full 6-skill workflow orchestration |
## Dual-Platform Skill Structure
@@ -96,7 +114,11 @@ XX-skill-name/
│ └── references/ # Documentation
├── desktop/ # Claude Desktop version
│ ├── SKILL.md # MCP-focused directive (YAML frontmatter)
│ ├── skill.yaml # Metadata (name, description, allowed-tools, license)
│ ├── SKILL.md # MCP-focused directive (content only, no frontmatter)
│ ├── tools/ # MCP tool documentation & configs
│ │ ├── {tool-name}.md # Tool usage docs (e.g., firecrawl.md)
│ │ └── {tool-name}.yaml # Tool config (optional)
│ ├── references/ # Guidance docs
│ └── examples/ # Usage examples
@@ -107,10 +129,28 @@ XX-skill-name/
| Aspect | `code/` | `desktop/` |
|--------|---------|------------|
| Directive | CLAUDE.md | SKILL.md (YAML) |
| Directive | CLAUDE.md | SKILL.md (content) |
| Metadata | In CLAUDE.md | skill.yaml |
| Tool docs | N/A | tools/ directory |
| Execution | Direct Bash/Python | MCP tools only |
| Scripts | Required | Reference only |
### skill.yaml Schema
```yaml
# Required fields
name: string # Skill identifier
description: string # Trigger description (can be multi-line with |)
# Optional fields
allowed-tools: # List of permitted MCP/Claude tools
- mcp__firecrawl__*
- Read
- Write
license: string # MIT, Internal-use Only, etc.
triggers: [] # Explicit trigger keywords
```
### Development Workflow
1. **Build Claude Code version first** - Full automation with scripts
@@ -166,6 +206,7 @@ our-claude-skills/
│ │ ├── 04-content-distiller/
│ │ ├── 05-quality-reviewer/
│ │ ├── 06-markdown-exporter/
│ │ ├── 07-pipeline-orchestrator/ # Full pipeline coordination
│ │ ├── shared/ # Config templates & MySQL schema
│ │ └── install.sh # Cross-machine installer
│ │