docs_skill

Overview

This skill generates clear, task-oriented technical documentation for projects, APIs, and code. It produces READMEs, API references, inline code docs (Python docstrings, JSDoc), user guides, and architecture overviews focused on what users need to accomplish. It emphasizes maintainability, scannability, and runnable examples.

How this skill works

I determine the documentation type, select an appropriate template, and adapt content to the project's goals and audience. For existing docs I run a quality audit against a checklist (runnable examples, valid links, accurate defaults) and patch gaps. Outputs include Markdown READMEs, endpoint-by-endpoint API docs, sample docstrings/JSDoc, architecture diagrams and exportable formats (HTML, PDF) when required.

When to use it

• Bootstrapping a new repository README or Quick Start
• Documenting REST or RPC APIs with parameter/response examples
• Adding or improving Python docstrings or JSDoc for libraries
• Creating architecture overviews and component data flows
• Auditing and fixing stale or incomplete documentation

Best practices

• Write for the reader’s skill level and primary goals, not internal implementation details
• Show the most common working example first, then explain theory
• Keep docs scannable: clear headers, tables for structured data, and short code blocks
• Scope docs to what will be actively maintained to avoid rot
• Validate that every code example runs and links resolve before delivery

Example use cases

• Generate a README with Quick Start, Features, Configuration, Usage, and Contributing sections
• Produce endpoint documentation for an API with parameter tables, response samples, and error codes
• Convert Python functions to Google-style docstrings with examples and exceptions documented
• Create an architecture overview listing components, responsibilities, data flow, and key decisions
• Audit an existing docs site and produce a prioritized fix list with updated samples