Menu
Generates professional API documentation, user guides, and architectural documentation from code and specifications.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | Documentation Builder |
| description | Generates professional API documentation, user guides, and architectural documentation from code and specifications. |
The Documentation Builder Skill transforms your codebase and specifications into professional, well-structured documentation. It analyzes your code, APIs, and architecture, then generates:
When to use this skill:
Tell the Documentation Builder what you need:
"Create API documentation for our REST endpoints including:
- Authentication (JWT)
- All CRUD operations
- Error responses
- Rate limiting info
- Code examples in Python and JavaScript"
The skill analyzes your code and generates:
Generated sections:
Quality standards:
Output formats:
Typical sections:
Getting Started
Core Concepts
How-To Guides
Advanced Topics
Reference
Typically includes:
Quality standards:
Input:
"Generate API documentation for our Flask REST API with:
- 8 endpoints (CRUD on two resources)
- JWT authentication
- JSON request/response
- Error handling
- Include Python and JavaScript examples"
Process:
1. Analyze Flask route definitions
2. Extract endpoint signatures
3. Parse request/response validation
4. Generate OpenAPI spec
5. Create markdown with examples
6. Format for multiple platforms
Output:
- api-documentation.md (complete reference)
- openapi.json (machine-readable spec)
- api-docs.html (hosted version)
- api-docs.pdf (downloadable version)
Input:
"Create a 'Getting Started' guide for developers setting up
our development environment"
Process:
1. Analyze setup scripts and configuration
2. Identify prerequisites and dependencies
3. Document installation steps
4. Create first-run tutorial
5. Add common gotchas and solutions
6. Include troubleshooting section
Output:
- getting-started.md
- Inline code examples
- Screenshots/diagrams
- Video tutorial outline
Input:
"Document our microservices architecture including:
- 5 services
- Kafka message queue
- PostgreSQL database
- Redis cache
- Load balancing approach"
Process:
1. Analyze service definitions
2. Map service boundaries
3. Document data flows
4. Create architecture diagram
5. Explain technology choices
6. Document deployment approach
7. Plan for scalability
Output:
- architecture.md (comprehensive overview)
- service-diagram.png (visual representation)
- data-flow.png (data pipeline)
- deployment.md (deployment guide)
Generate documentation in multiple formats simultaneously:
Auto-generate code examples in multiple languages:
Creates diagrams for complex systems:
Keeps documentation in sync:
Provide context so skill understands your system:
"Create API docs for our payment processing endpoint
that uses Stripe, includes webhook handling, and PCI compliance notes"
Instead of:
"Document the payment API"
Specify what examples would be most useful:
"Include examples showing:
- Successful payment processing flow
- Handling failed payments
- Setting up webhook handlers
- Common error scenarios"
Tell the skill who will read the docs:
"Write for mid-level Python developers who know requests
but may be new to async patterns"
Different docs serve different purposes:
Regenerate when significant changes occur:
Specify which languages to include:
{
"example_languages": ["python", "javascript", "typescript"],
"include_curl": true,
"include_postman": true
}
Choose which sections to include:
{
"sections": [
"overview",
"authentication",
"endpoints",
"errors",
"examples",
"webhooks",
"rate_limiting"
]
}
Customize appearance:
{
"brand": {
"company_name": "MyCompany",
"logo_url": "https://...",
"primary_color": "#007AFF"
}
}
Specify destination formats:
{
"outputs": [
"markdown",
"openapi.json",
"html",
"pdf",
"docusaurus"
]
}
Estimated tokens:
✅ Documentation is complete when:
Cause: Code structure not analyzed properly Solution: Provide explicit endpoint list
Cause: Context not specific enough Solution: Provide actual code or detailed descriptions
Cause: Complex markdown structure Solution: Specify desired output format (Markdown, HTML, etc.)
For technical implementation details, see REFERENCE.md