API Documentation Generator

Automatically extract API routes from your codebase and generate comprehensive OpenAPI 3.0 specifications with proper schemas, examples, and descriptions.

Quick Start

When a user asks for API documentation, follow this checklist:

API Documentation Progress:
- [ ] Step 1: Detect API framework and analyze routes
- [ ] Step 2: Extract endpoint definitions
- [ ] Step 3: Infer request/response schemas
- [ ] Step 4: Generate OpenAPI specification
- [ ] Step 5: Create markdown documentation
- [ ] Step 6: Add examples and descriptions
- [ ] Step 7: Validate and output

---

Workflow

Step 1: Detect Framework and Analyze Routes

Run the route analyzer to detect the API framework and find all endpoints:

python .claude/skills/api-docs-generator/scripts/analyze_routes.py .

Supported Frameworks:

• Express.js / Fastify / Hono / NestJS (Node.js)
• FastAPI / Flask / Django REST (Python)
• Gin / Echo / Chi (Go)
• Next.js API Routes / App Router

Output includes:

• Framework detected
• Base URL/prefix
• List of all routes with methods
• Middleware detected (auth, validation)

---

Step 2: Extract Endpoint Definitions

For each detected endpoint, extract:

Property	Source
Path	Route definition
Method	GET, POST, PUT, PATCH, DELETE
Parameters	Path params, query params
Request body	Validation schema, TypeScript types
Response	Return statements, response calls
Auth	Middleware, decorators
Tags	File/folder structure

Express Example:

// Route: POST /api/users
// Extract: path params, body schema, response type
router.post("/", validateRequest(createUserSchema), userController.create);

---

Step 3: Infer Request/Response Schemas

Extract schemas from validation libraries and TypeScript types:

From Zod

const createUserSchema = z.object({
  email: z.string().email(),
  name: z.string().min(2).max(100),
  role: z.enum(["admin", "user"]),
});
// → Infer OpenAPI schema with types, formats, constraints

From TypeScript Interfaces

interface UserResponse {
  id: string;
  email: string;
  createdAt: Date;
}
// → Convert to OpenAPI schema

From Pydantic (Python)

class UserCreate(BaseModel):
    email: EmailStr
    name: str = Field(..., min_length=2, max_length=100)
# → Convert to OpenAPI schema

Schema Mapping:

Source Type	OpenAPI Type	Format
string	string	-
number	number	-
boolean	boolean	-
Date	string	date-time
email	string	email
uuid	string	uuid
url	string	uri
int	integer	int32
array	array	-
object	object	-

---

Step 4: Generate OpenAPI Specification

Create the OpenAPI 3.0 YAML/JSON file:

openapi: 3.0.3
info:
  title: { { API_NAME } }
  description: { { API_DESCRIPTION } }
  version: { { VERSION } }
  contact:
    name: API Support
    email: support@example.com

servers:
  - url: http://localhost:3000
    description: Development
  - url: https://api.example.com
    description: Production

paths:
  /api/users:
    get:
      summary: List users
      operationId: listUsers
      tags:
        - Users
      parameters:
        - $ref: "#/components/parameters/PageParam"
        - $ref: "#/components/parameters/LimitParam"
      responses:
        "200":
          description: Successful response
          content:
            application/json:
              schema:
                type: array
                items:
                  $ref: "#/components/schemas/User"
    post:
      summary: Create user
      operationId: createUser
      tags:
        - Users
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: "#/components/schemas/CreateUserInput"
      responses:
        "201":
          description: Created
          content:
            application/json:
              schema:
                $ref: "#/components/schemas/User"
        "400":
          $ref: "#/components/responses/BadRequest"

components:
  schemas:
    User:
      type: object
      properties:
        id:
          type: string
          format: uuid
        email:
          type: string
          format: email
        name:
          type: string
        createdAt:
          type: string
          format: date-time
      required:
        - id
        - email
        - name
        - createdAt

  parameters:
    PageParam:
      name: page
      in: query
      schema:
        type: integer
        default: 1
        minimum: 1
    LimitParam:
      name: limit
      in: query
      schema:
        type: integer
        default: 20
        minimum: 1
        maximum: 100

  responses:
    BadRequest:
      description: Invalid request
      content:
        application/json:
          schema:
            $ref: "#/components/schemas/Error"

  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      bearerFormat: JWT

---

Step 5: Create Markdown Documentation

Generate human-readable API documentation:

python .claude/skills/api-docs-generator/scripts/generate_markdown.py openapi.yaml --output docs/API.md

Output Structure:

# API Documentation

## Authentication

Bearer token required for protected endpoints.

## Endpoints

### Users

#### List Users

`GET /api/users`

**Parameters:**
| Name | Type | Required | Description |
|------|------|----------|-------------|
| page | integer | No | Page number (default: 1) |
| limit | integer | No | Items per page (default: 20) |

**Response:**

```json
[
  {
    "id": "abc123",
    "email": "user@example.com",
    "name": "John Doe",
    "createdAt": "2024-01-15T10:30:00Z"
  }
]
```

Create User

POST /api/users

Request Body:

{
  "email": "user@example.com",
  "name": "John Doe",
  "role": "user"
}

...

---

### Step 6: Add Examples and Descriptions

Enhance documentation with:

1. **Request/Response Examples**
   - Generate realistic sample data
   - Include edge cases
   - Show error responses

2. **Descriptions**
   - Summarize each endpoint's purpose
   - Document business rules
   - Note rate limits and permissions

3. **Code Snippets**
   - cURL examples
   - JavaScript/fetch examples
   - Python requests examples

---

### Step 7: Validate and Output

Validate the generated OpenAPI spec:

```bash
python .claude/skills/api-docs-generator/scripts/validate_openapi.py openapi.yaml

Checks:

• Valid OpenAPI 3.0 syntax
• All $ref references resolve
• Required fields present
• Schema consistency

Output Files:

docs/
├── openapi.yaml          # OpenAPI 3.0 spec
├── openapi.json          # JSON version
├── API.md                # Markdown docs
└── postman_collection.json  # Postman import

---

Configuration

Create .claude/api-docs-config.yaml to customize:

info:
  title: My API
  version: 1.0.0
  description: Backend API for MyApp

servers:
  - url: http://localhost:3000
    description: Development
  - url: https://api.myapp.com
    description: Production

defaults:
  auth: BearerAuth
  content_type: application/json
  tag_from: folder # folder, file, or manual

output:
  format: yaml # yaml or json
  path: docs/openapi.yaml
  markdown: docs/API.md
  postman: true

ignore:
  - /health
  - /metrics
  - /internal/*

---

Framework-Specific Extraction

Express.js

// Patterns detected:
router.get("/users", handler); // Basic route
router.post("/users", validate, handler); // With middleware
app.use("/api", apiRouter); // Nested routers

FastAPI

# Patterns detected:
@router.get("/users", response_model=List[User])
@router.post("/users", status_code=201)
# + Pydantic models for schemas

Next.js App Router

// app/api/users/route.ts
export async function GET(request: Request) {}
export async function POST(request: Request) {}
// → Infer routes from file structure

NestJS

// Patterns detected:
@Controller('users')
@Get()
@Post()
@ApiOperation({ summary: 'List users' })  // Already has OpenAPI decorators

---

Examples

Example 1: Express API

User asks: "Generate OpenAPI docs for my API"

1. Run analyze_routes.py → Detects Express
2. Find routes in src/routes/*.ts
3. Extract Zod schemas from src/schemas/
4. Generate openapi.yaml
5. Create API.md

Example 2: FastAPI

User asks: "Create Swagger documentation"

1. Detect FastAPI
2. Parse routers with decorators
3. Extract Pydantic models
4. Generate spec (FastAPI has built-in support, enhance it)

Example 3: Next.js

User asks: "Document my Next.js API routes"

1. Scan app/api/** or pages/api/**
2. Infer routes from file structure
3. Parse request/response types
4. Generate documentation

---

Tips for Best Results

1. Use validation libraries - Zod, Pydantic, etc. provide schema info
2. Add TypeScript types - Help infer request/response shapes
3. Organize routes by resource - Creates better tags automatically
4. Include JSDoc/docstrings - Used for descriptions
5. Use consistent patterns - Easier to detect and extract

---

Related Skills

• fullstack-feature-generator - Generate APIs with built-in docs
• codebase-onboarding - Understand existing APIs first