Removals (duplicate/stale): - docs/DOCKER_GUIDE.md: 80% overlap with DOCKER_DEPLOYMENT.md - docs/KUBERNETES_GUIDE.md: 70% overlap with KUBERNETES_DEPLOYMENT.md - docs/strategy/TASK19_COMPLETE.md: stale task tracking - docs/strategy/TASK20_COMPLETE.md: stale task tracking - docs/strategy/TASK21_COMPLETE.md: stale task tracking - docs/strategy/WEEK2_COMPLETE.md: stale progress report Updates (version/counts): - docs/FAQ.md: v2.7.0 → v3.1.0-dev, 18 MCP tools → 26, 4 platforms → 16+ - docs/QUICK_REFERENCE.md: 18 MCP tools → 26, 1200+ tests → 1,880+, footer updated - docs/features/BOOTSTRAP_SKILL.md: v2.7.0 → v3.1.0-dev header and footer Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
17 KiB
Bootstrap Skill - Self-Hosting (v3.1.0-dev)
Version: 3.1.0-dev Feature: Bootstrap Skill (Dogfooding) Status: ✅ Production Ready Last Updated: 2026-02-18
Overview
The Bootstrap Skill feature allows Skill Seekers to analyze itself and generate a Claude Code skill containing its own documentation, API reference, code patterns, and usage examples. This is the ultimate form of "dogfooding" - using the tool to document itself.
What You Get:
- Complete Skill Seekers documentation as a Claude Code skill
- CLI command reference with examples
- Auto-generated API documentation from codebase
- Design pattern detection from source code
- Test example extraction for learning
- Installation into Claude Code for instant access
Use Cases:
- Learn Skill Seekers by having it explain itself to Claude
- Quick reference for CLI commands while working
- API documentation for programmatic usage
- Code pattern examples from the source
- Self-documenting development workflow
Quick Start
One-Command Installation
# Generate and install the bootstrap skill
./scripts/bootstrap_skill.sh
This script will:
- ✅ Analyze the Skill Seekers codebase (C3.x features)
- ✅ Merge handcrafted header with auto-generated content
- ✅ Validate YAML frontmatter and structure
- ✅ Create
output/skill-seekers/directory - ✅ Install to Claude Code (optional)
Time: ~2-5 minutes (depending on analysis depth)
Manual Installation
# 1. Run codebase analysis
skill-seekers codebase \
--directory . \
--output output/skill-seekers \
--name skill-seekers
# 2. Merge with custom header (optional)
cat scripts/skill_header.md output/skill-seekers/SKILL.md > output/skill-seekers/SKILL_MERGED.md
mv output/skill-seekers/SKILL_MERGED.md output/skill-seekers/SKILL.md
# 3. Install to Claude Code
skill-seekers install-agent \
--skill-dir output/skill-seekers \
--agent-dir ~/.claude/skills/skill-seekers
How It Works
Architecture
The bootstrap skill combines three components:
┌─────────────────────────────────────────────────────────┐
│ Bootstrap Skill Architecture │
├─────────────────────────────────────────────────────────┤
│ │
│ 1. Handcrafted Header (scripts/skill_header.md) │
│ ├── YAML frontmatter │
│ ├── Installation instructions │
│ ├── Quick start guide │
│ └── Core concepts │
│ │
│ 2. Auto-Generated Content (codebase_scraper.py) │
│ ├── C3.1: Design pattern detection │
│ ├── C3.2: Test example extraction │
│ ├── C3.3: How-to guide generation │
│ ├── C3.4: Configuration extraction │
│ ├── C3.5: Architectural overview │
│ ├── C3.7: Architectural pattern detection │
│ ├── C3.8: API reference + dependency graphs │
│ └── Code analysis (9 languages) │
│ │
│ 3. Validation System (frontmatter detection) │
│ ├── YAML frontmatter check │
│ ├── Required field validation │
│ └── Structure verification │
│ │
└─────────────────────────────────────────────────────────┘
Step 1: Codebase Analysis
The codebase_scraper.py module analyzes the Skill Seekers source code:
skill-seekers codebase --directory . --output output/skill-seekers
What Gets Analyzed:
- Python source files (
src/skill_seekers/**/*.py) - Test files (
tests/**/*.py) - Configuration files (
configs/*.json) - Documentation (
docs/**/*.md,README.md, etc.)
C3.x Features Applied:
- C3.1: Detects design patterns (Strategy, Factory, Singleton, etc.)
- C3.2: Extracts test examples showing real usage
- C3.3: Generates how-to guides from test workflows
- C3.4: Extracts configuration patterns (CLI args, env vars)
- C3.5: Creates architectural overview of the codebase
- C3.7: Detects architectural patterns (MVC, Repository, etc.)
- C3.8: Builds API reference and dependency graphs
Step 2: Header Combination
The bootstrap script merges a handcrafted header with auto-generated content:
# scripts/bootstrap_skill.sh does this:
cat scripts/skill_header.md output/skill-seekers/SKILL.md > merged.md
Why Two Parts?
- Header: Curated introduction, installation steps, core concepts
- Auto-generated: Always up-to-date code patterns, examples, API docs
Header Structure (scripts/skill_header.md):
---
name: skill-seekers
version: 2.7.0
description: |
Documentation-to-AI skill conversion tool. Use when working with
Skill Seekers codebase, CLI commands, or API integration.
tags: [documentation, scraping, ai-skills, mcp]
---
# Skill Seekers - Documentation to AI Skills
## Installation
...
## Quick Start
...
## Core Concepts
...
<!-- AUTO-GENERATED CONTENT STARTS HERE -->
Step 3: Validation
The bootstrap script validates the final skill:
# Check for YAML frontmatter
if ! grep -q "^---$" output/skill-seekers/SKILL.md; then
echo "❌ Missing YAML frontmatter"
exit 1
fi
# Validate required fields
python -c "
import yaml
with open('output/skill-seekers/SKILL.md') as f:
content = f.read()
frontmatter = yaml.safe_load(content.split('---')[1])
required = ['name', 'version', 'description']
for field in required:
assert field in frontmatter, f'Missing {field}'
"
Validated Fields:
- ✅
name- Skill name - ✅
version- Version number - ✅
description- When to use this skill - ✅
tags- Categorization tags - ✅ Proper YAML syntax
- ✅ Content structure
Step 4: Output
The final skill is created in output/skill-seekers/:
output/skill-seekers/
├── SKILL.md # Main skill file (300-500 lines)
├── references/ # Detailed references
│ ├── api_reference/ # API documentation
│ │ ├── doc_scraper.md
│ │ ├── github_scraper.md
│ │ └── ...
│ ├── patterns/ # Design patterns detected
│ │ ├── strategy_pattern.md
│ │ ├── factory_pattern.md
│ │ └── ...
│ ├── test_examples/ # Usage examples from tests
│ │ ├── scraping_examples.md
│ │ ├── packaging_examples.md
│ │ └── ...
│ └── how_to_guides/ # Generated guides
│ ├── how_to_scrape_docs.md
│ ├── how_to_package_skills.md
│ └── ...
└── metadata.json # Skill metadata
Advanced Usage
Customizing the Header
Edit scripts/skill_header.md to customize the introduction:
---
name: skill-seekers
version: 2.7.0
description: |
YOUR CUSTOM DESCRIPTION HERE
tags: [your, custom, tags]
custom_field: your_value
---
# Your Custom Title
Your custom introduction...
<!-- AUTO-GENERATED CONTENT STARTS HERE -->
Guidelines:
- Keep frontmatter in YAML format
- Include required fields:
name,version,description - Add custom fields as needed
- Marker comment preserves auto-generated content location
Validation Options
The bootstrap script supports custom validation rules:
# scripts/bootstrap_skill.sh (excerpt)
# Custom validation function
validate_skill() {
local skill_file=$1
# Check frontmatter
if ! has_frontmatter "$skill_file"; then
echo "❌ Missing frontmatter"
return 1
fi
# Check required fields
if ! has_required_fields "$skill_file"; then
echo "❌ Missing required fields"
return 1
fi
# Check content structure
if ! has_proper_structure "$skill_file"; then
echo "❌ Invalid structure"
return 1
fi
echo "✅ Validation passed"
return 0
}
Custom Validation:
- Add your own validation functions
- Check for custom frontmatter fields
- Validate content structure
- Enforce your own standards
CI/CD Integration
Automate bootstrap skill generation in your CI/CD pipeline:
# .github/workflows/bootstrap-skill.yml
name: Generate Bootstrap Skill
on:
push:
branches: [main, development]
schedule:
- cron: '0 0 * * 0' # Weekly on Sunday
jobs:
bootstrap:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- uses: actions/setup-python@v4
with:
python-version: '3.11'
- name: Install Skill Seekers
run: pip install -e .
- name: Generate Bootstrap Skill
run: ./scripts/bootstrap_skill.sh
- name: Upload Artifact
uses: actions/upload-artifact@v3
with:
name: bootstrap-skill
path: output/skill-seekers/
- name: Commit to Repository (optional)
run: |
git config user.name "GitHub Actions"
git config user.email "actions@github.com"
git add output/skill-seekers/
git commit -m "chore: Update bootstrap skill [skip ci]"
git push
Troubleshooting
Common Issues
1. Missing YAML Frontmatter
Error:
❌ Missing YAML frontmatter in output/skill-seekers/SKILL.md
Solution:
# Check if scripts/skill_header.md has frontmatter
cat scripts/skill_header.md | head -10
# Should start with:
# ---
# name: skill-seekers
# version: 2.7.0
# ...
# ---
2. Validation Failure
Error:
❌ Missing required fields in frontmatter
Solution:
# Check frontmatter fields
python -c "
import yaml
with open('output/skill-seekers/SKILL.md') as f:
content = f.read()
fm = yaml.safe_load(content.split('---')[1])
print('Fields:', list(fm.keys()))
"
# Ensure: name, version, description are present
3. Codebase Analysis Fails
Error:
❌ skill-seekers codebase failed with exit code 1
Solution:
# Run analysis manually to see error
skill-seekers codebase --directory . --output output/test
# Common causes:
# - Missing dependencies: pip install -e ".[all-llms]"
# - Invalid Python files: check syntax errors
# - Permission issues: check file permissions
4. Header Merge Issues
Error:
Auto-generated content marker not found
Solution:
# Ensure marker exists in header
grep "AUTO-GENERATED CONTENT STARTS HERE" scripts/skill_header.md
# If missing, add it:
echo "<!-- AUTO-GENERATED CONTENT STARTS HERE -->" >> scripts/skill_header.md
Debugging
Enable verbose output for debugging:
# Run with bash -x for debugging
bash -x ./scripts/bootstrap_skill.sh
# Or add debug statements
set -x # Enable debugging
./scripts/bootstrap_skill.sh
set +x # Disable debugging
Debug Checklist:
- ✅ Skill Seekers installed:
skill-seekers --version - ✅ Python 3.10+:
python --version - ✅ Dependencies installed:
pip install -e ".[all-llms]" - ✅ Header file exists:
ls scripts/skill_header.md - ✅ Output directory writable:
touch output/test && rm output/test
Testing
Running Tests
The bootstrap skill feature has comprehensive test coverage:
# Unit tests for bootstrap logic
pytest tests/test_bootstrap_skill.py -v
# End-to-end tests
pytest tests/test_bootstrap_skill_e2e.py -v
# Full test suite (10 tests for bootstrap feature)
pytest tests/test_bootstrap*.py -v
Test Coverage:
- ✅ Header parsing and validation
- ✅ Frontmatter detection
- ✅ Required field validation
- ✅ Content merging
- ✅ Output directory structure
- ✅ Codebase analysis integration
- ✅ Error handling
- ✅ Edge cases (missing files, invalid YAML, etc.)
E2E Test Example
def test_bootstrap_skill_e2e(tmp_path):
"""Test complete bootstrap skill workflow."""
# Setup
output_dir = tmp_path / "skill-seekers"
header_file = "scripts/skill_header.md"
# Run bootstrap
result = subprocess.run(
["./scripts/bootstrap_skill.sh"],
capture_output=True,
text=True
)
# Verify
assert result.returncode == 0
assert (output_dir / "SKILL.md").exists()
assert has_valid_frontmatter(output_dir / "SKILL.md")
assert has_required_fields(output_dir / "SKILL.md")
Test Coverage Report
# Run with coverage
pytest tests/test_bootstrap*.py --cov=scripts --cov-report=html
# View report
open htmlcov/index.html
Examples
Example 1: Basic Bootstrap
# Generate bootstrap skill
./scripts/bootstrap_skill.sh
# Output:
# ✅ Analyzing Skill Seekers codebase...
# ✅ Detected 15 design patterns
# ✅ Extracted 45 test examples
# ✅ Generated 12 how-to guides
# ✅ Merging with header...
# ✅ Validating skill...
# ✅ Bootstrap skill created: output/skill-seekers/SKILL.md
Example 2: Custom Analysis Depth
# Run with basic analysis (faster)
skill-seekers codebase \
--directory . \
--output output/skill-seekers \
--skip-patterns \
--skip-how-to-guides
# Then merge with header
cat scripts/skill_header.md output/skill-seekers/SKILL.md > merged.md
Example 3: Install to Claude Code
# Generate and install
./scripts/bootstrap_skill.sh
# Install to Claude Code
skill-seekers install-agent \
--skill-dir output/skill-seekers \
--agent-dir ~/.claude/skills/skill-seekers
# Now use in Claude Code:
# "Use the skill-seekers skill to explain how to scrape documentation"
Example 4: Programmatic Usage
from skill_seekers.cli.codebase_scraper import scrape_codebase
from skill_seekers.cli.install_agent import install_to_agent
# 1. Analyze codebase
result = scrape_codebase(
directory='.',
output_dir='output/skill-seekers',
name='skill-seekers',
enable_patterns=True,
enable_how_to_guides=True
)
print(f"Skill created: {result['skill_path']}")
# 2. Merge with header
with open('scripts/skill_header.md') as f:
header = f.read()
with open(result['skill_path']) as f:
content = f.read()
merged = header + "\n\n<!-- AUTO-GENERATED -->\n\n" + content
with open(result['skill_path'], 'w') as f:
f.write(merged)
# 3. Install to Claude Code
install_to_agent(
skill_dir='output/skill-seekers',
agent_dir='~/.claude/skills/skill-seekers'
)
print("✅ Bootstrap skill installed to Claude Code!")
Performance Characteristics
| Operation | Time | Notes |
|---|---|---|
| Codebase analysis | 1-3 min | With all C3.x features |
| Header merging | <1 sec | Simple concatenation |
| Validation | <1 sec | YAML parsing + checks |
| Installation | <1 sec | Copy to agent directory |
| Total | 2-5 min | End-to-end bootstrap |
Analysis Breakdown:
- Pattern detection (C3.1): ~30 sec
- Test extraction (C3.2): ~20 sec
- How-to guides (C3.3): ~40 sec
- Config extraction (C3.4): ~10 sec
- Architecture overview (C3.5): ~30 sec
- Arch pattern detection (C3.7): ~20 sec
- API reference (C3.8): ~30 sec
Best Practices
1. Keep Header Minimal
The header should provide context and quick start, not duplicate auto-generated content:
---
name: skill-seekers
version: 2.7.0
description: Brief description
---
# Quick Introduction
Essential information only.
<!-- AUTO-GENERATED CONTENT STARTS HERE -->
2. Regenerate Regularly
Keep the bootstrap skill up-to-date with codebase changes:
# Weekly or on major changes
./scripts/bootstrap_skill.sh
# Or automate in CI/CD
3. Version Header with Code
Keep scripts/skill_header.md in version control:
git add scripts/skill_header.md
git commit -m "docs: Update bootstrap skill header"
4. Validate Before Committing
Always validate the generated skill:
# Run validation
python -c "
import yaml
with open('output/skill-seekers/SKILL.md') as f:
content = f.read()
assert '---' in content, 'Missing frontmatter'
fm = yaml.safe_load(content.split('---')[1])
assert 'name' in fm
assert 'version' in fm
"
echo "✅ Validation passed"
Related Features
- Codebase Scraping - Analyze local codebases
- C3.x Features - Pattern detection and analysis
- Install Agent - Install skills to Claude Code
- API Reference - Programmatic usage
Changelog
v2.7.0 (2026-01-18)
- ✅ Bootstrap skill feature introduced
- ✅ Dynamic frontmatter detection (not hardcoded)
- ✅ Comprehensive validation system
- ✅ CI/CD integration examples
- ✅ 10 unit tests + 8-12 E2E tests
Version: 3.1.0-dev Last Updated: 2026-02-18 Status: ✅ Production Ready