write-documentation

Generate accurate API references by reading source code signatures, schemas, and defaults to produce documentation derived directly from the actual implementation rather than memory. This ensures fidelity with current behavior and links to canonical sources.

Keep README updates and changelogs in sync by deriving documentation content directly from source code artifacts and wiring those updates into the correct project index and changelog automatically during the documentation authoring workflow.

Yes, you can write documentation for code-oriented projects across languages and stacks. The process adapts to audience requirements by reading signatures and schemas from the source, ensuring the output fits the specific project context.

Yes, code-derived documentation integrates with the architectural decision record (ADR) workflow. When documentation generation surfaces relevant architectural decisions, it links those decisions to the ADR workflow for proper project tracking.

Avoid documentation duplication by linking inline docs and API references to canonical sources rather than copying content. The generation process reads existing references and source artifacts to ensure a single source of truth.

Automated documentation generation is not suitable for projects lacking stable source code signatures or schemas to read. It requires reading actual implementation details, so it fails when the codebase is missing or behavior is entirely undocumented.