Generate comprehensive API documentation from codebases

API Documentation Generator turns an API implementation into documentation for developers integrating with it. It supports REST, GraphQL, and WebSocket APIs, with attention to endpoint descriptions, request and response schemas, authentication, error handling, and usage examples. The process starts with the actual source revision, routes, middleware, and schemas so the published explanation reflects implemented behavior.

The skill distinguishes verified facts from missing or inferred behavior. It calls for clarification rather than inventing rate limits, password policies, token expiry, or an exhaustive error catalog. Examples use synthetic data and secret placeholders. Public account creation must not suggest accepting a client-supplied administrator role unless an authorized endpoint explicitly validates that operation.

Use it for a new API, an existing API with incomplete documentation, developer onboarding, or documentation intended for external users. It also supports creating OpenAPI or Swagger specifications and organizing material into a quickstart, resource-based endpoint reference, data models, error handling, versioning, and changelog.

Do not treat illustrative payloads as evidence of production behavior. Confirm the target environment and permissions before running examples. Interactive Try It operations belong in an approved sandbox, and a generated specification needs complete references and validation before use.

Inputs include API code, endpoint definitions, request validation, authentication middleware, and verified examples. For each endpoint, the documentation identifies the HTTP method, path, purpose, authentication requirements, and applicable limits. Parameters are separated into path, query, header, and body inputs, with required or optional status, types, formats, and validation rules.

Outputs describe response status codes, bodies, headers, and relevant failure scenarios. Error guidance pairs error formats with troubleshooting examples instead of leaving integration failures unexplained. Usage material can cover pagination, filtering, sorting, and common workflows when the implementation supports them.

Code examples can use curl, JavaScript or TypeScript with fetch or axios, and Python requests. An authentication header example is:

Authorization: Bearer YOUR_TOKEN

The skill also addresses keeping documentation synchronized: review it when the API changes, use annotations or OpenAPI where appropriate, and check examples against API behavior. Complete working examples are preferred over fragments that omit required context. Constraints, version information, and related endpoints help readers interpret each operation accurately.

OpenAPI and Swagger support structured API specifications. Postman collections and interactive examples provide additional ways to explore documented operations. These outputs still require valid references, realistic sample responses, and execution checks in the approved environment. Authentication documentation can explain login, token use, and refresh flows when those behaviors are verified rather than assuming that an example token lifetime is a product guarantee.