Generate Automated Code Documentation

A skill for generating living documentation - API docs, architecture diagrams, user guides, and technical references - directly from a codebase using AI-powered analysis, rather than writing docs by hand. It extracts information from code, configuration files, and comments, then produces docs with consistent terminology and structure across doc types, plus automation (linting, CI checks) to keep documentation synchronized with the code it describes over time.

Use when generating API, architecture, or user documentation from an existing codebase, building a documentation pipeline or automation, or standardizing docs across a repository. Do not use it when the project has no codebase or source of truth to extract from, when only an ad-hoc explanation is needed rather than durable documentation, or when there is no access to the code or requirements to ground the docs in.

Input is the target codebase plus a free-form requirements argument describing what documentation is needed. Output is a documentation plan and the artifacts themselves, plus file paths and tooling configuration for any doc-generation automation, and an explicit list of assumptions, gaps, and follow-up tasks. For detailed examples and templates, it points to resources/implementation-playbook.md.

Designed to plug into CI pipelines and linting tooling to keep generated documentation synchronized with code changes over time; the specific stack is left to the implementation playbook rather than fixed here.

Teams that need documentation to stay current with a changing codebase - API references, architecture docs, or onboarding guides - rather than a one-off written explanation.