Author and verify documentation, diagrams, and programmatic office documents. Use for docs verification, runbooks, ADRs, guides, Mermaid diagrams as code, ASCII and terminal figures, code snippet images, PDF/DOCX generation, and OpenAPI reference. Trigger on 'document this', 'write a runbook', 'draw a diagram', 'generate a PDF', 'are these docs still true'.
Records decisions and documentation. Use when making architectural decisions, changing public APIs, shipping features, or when you need to record context that future engineers and agents will need to understand the codebase.
Enforce minimum documentation standard across uFawkes* repos. Use when auditing a repo before release, onboarding a new repo to the suite, or implementing roadmap item 0.6 (cross-repo README links). Directly supports DORA AI Capability 3 and the 2023 DORA finding that documentation quality amplifies all technical capabilities.
Expert technical documentation architect specializing in developer experience, API portals, interactive documentation, and documentation-as-code strategies.
Documents code the way the language's own toolchain expects: the native doc format per language, contract-first doc comments on every public symbol, module-level overviews, examples the test suite runs so they cannot drift, and inline comments that explain why rather than what. Use whenever code is written, added, refactored, or reviewed, whenever a public API changes, and whenever asked to write, fix, or review comments, docstrings, doc comments, or API reference documentation.
Durable Documentation
**Status:** experimental.
Governs durable, edited repository knowledge. Unlike the append-only log, documentation is actively edited as plans and decisions evolve.
## Directory Structure
A project managed by Agent
Write and maintain the /docs knowledge base. Apply when adding or changing a page under /docs, or when a code change makes existing documentation wrong.
Use before you edit a README.md, a crate README, anything under docs/, an architecture doc, or a protocol doc, when a behaviour change could leave a doc stale, when the code and a protocol doc disagree, when you decide whether a sentence belongs in a protocol doc, a README, a plan, a test, or the code, and when a change or a review option conflicts with a protocol doc.
Write and maintain the /docs knowledge base. Apply when adding or changing a page under /docs, or when a code change makes existing documentation wrong.
Write and maintain the /docs knowledge base. Apply when adding or changing a page under /docs, or when a code change makes existing documentation wrong.
Maintain project documentation: README (Standard README spec), Architecture Decision Records, Diátaxis technical docs, and Keep a Changelog CHANGELOG.md. Use when writing or auditing README, ADR, docs site content, or changelog entries, or when the user mentions documentation structure or doc maintenance.
文档更新与维护。检查并更新README.md、CHANGELOG.md、SECURITY.md、部署文档和.gitignore模板。Use when the user asks to update documentation, sync changelog, or prepare release docs.
Documentation standards and required templates for this project. Use this skill whenever producing any project document — feature plans, handoffs, release notes, capability documentation, production release records, or incident reports. Also trigger it when a plan, release, deployment, or incident occurs without documentation, because in this project undocumented work is unfinished work.
Writes or reviews developer documentation in the project style with SEO guidance. Use when adding, changing, or removing content in a docs page, a README, an option or API description, a guide, a FAQ, or a blog post, even when the request only says "document this".
Plain text files in a repository that tell a coding agent how the project works: commands to run, conventions to follow and things to avoid. CLAUDE.md, AGENTS.md, cursor rules and skills are the common kinds.
CLAUDE.md or AGENTS.md?
CLAUDE.md is read by Claude Code. AGENTS.md is an open format that Codex, Cursor and other agents read. Many projects keep one and point the other at it.
What is a skill?
A folder with a SKILL.md that describes one capability, such as filling PDFs or reviewing code. The agent loads it only when the task calls for it.
Can I search my own team's files too?
Your agents already can, over MCP, limited to the files you're allowed to read. Searching them from this page is coming.