reconciling-documentation
Sweeps a repository's documentation for claims that disagree with its code and tests, decides which side is wrong for each, and fixes it after approval. Use when the user asks to reconcile, audit, or fix documentation that has drifted from the code.
- License 许可
- MIT
- 最近更新
- 2026-11-15
- Stars
- ★ 2
- Forks
- 0
01 前言 Front Matter
- name
- reconciling-documentation
- metadata
- { "internal": true }
- description
- Sweeps a repository's documentation for claims that disagree with its code and tests, decides which side is wrong for each, and fixes it after approval. Use when the user asks to reconcile, audit, or fix documentation that has drifted from the code.
- disable-model-invocation
- true
02 资源文件 2
- 参考文档agents/openai.yaml260 B
- 参考文档shared/docs-vs-code.md1.3 KB
03 SKILL.md 正文
Reconciling documentation
Tests check the code against what its authors meant. Nothing checks the code against what the docs say. This skill closes that gap for the scope the user names; with none named, the scope is the README, AGENTS.md, CLAUDE.md, and docs/.
Every verdict follows the docs and code rules.
- List claims. In each document in scope, record every claim a reader could check: a command, a path, a default, an option, a function's behavior. Skip opinions and plans. Keep each claim's
file:line. - Check each claim against the code and the tests. Record the code's
file:lineand the test that pins the behavior, or "no test". Run a documented command only when it is read-only. - Classify each mismatch as a code bug, a stale doc, or unsettled. For an unsettled one, gather intent evidence before asking: callers that depend on one behavior, the history of both sides (
git log -L,git blame), and any spec or issue. A side someone changed later on purpose usually states the intent. - Present the plan, then stop. One row per mismatch: claim, doc line, code line, pinning test, verdict, planned fix. Ask the user to approve the plan and to settle each unsettled item. Change nothing before the answer.
- Fix the approved rows.
- Code bug: write a test that fails on the current code, run it to see it fail, change the code, and run it to see it pass.
- Stale doc: edit the document only; leave the code and tests as they are.
- Verify. Run the project's full test command once after the last fix. Report each row as fixed, skipped, or still open, with the test command's result.
04 同仓库更多 Skills
writing-system-prompts
Writes and reviews system prompts for AI tools, agents, chatbots, and LLM products using a fixed section order and a quality checklist. Use when the user says "write a system prompt", or when drafting a system prompt or auditing one for missing stop criteria, tool rules, constraints, or examples. Not for compressing an existing prompt; use improving-prompts.
writing-prose
Writes and edits prose for people in a plain, conversational, short house style with no em dashes. Use when the user says "write this up", or when drafting or revising documentation, READMEs, explanations, blog posts, or other human-facing prose. Not for prompts written for agents; use writing-system-prompts or composing-agent-prompts.
writing-acceptance-criteria
Writes or rewrites acceptance criteria until each is provable: one observable outcome with a named check that can fail. Use when the user says "write acceptance criteria", or asks to write, rewrite, or tighten acceptance criteria or a definition of done. Not for proving work meets them; use proving-claims.
summarizing-shipped-work
Summarizes the user's merged GitHub PRs for a timeframe into a grouped, linked shipped-work report with highlights, and offers to update a brag document. Use when the user says "what shipped?", or when asked what the user shipped (this week, last month, a quarter, since a date), for a status update, or to update a brag doc.
running-retros
Mines a session transcript, past agent sessions, or named sources such as PR review comments for durable learnings, then proposes approval-gated edits to skills, AGENTS.md, or docs and tracker issues. Use when the user says "retro", or when the user explicitly asks for a retrospective. Never infer from session end or friction.
learning-from-mistakes
Codifies the latest correction or mistake in the conversation as a short imperative rule in the project or global CLAUDE.md or AGENTS.md, after the user approves it. Use when the user says "learn from this mistake", or when the user asks to learn from a mistake, remember a lesson for next time, add a rule to CLAUDE.md, or not repeat a behavior. Not for a whole session; use running-retros.