All products

Audit project documentation with AI

Follow real reader tasks through the project, detect outdated destinations and report evidence without reorganising files during the audit.

Audit project documentation with AI by following a task from its real starting point. A migration link can open perfectly and still lead a new developer to instructions for the wrong release. The docs-audit skill checks whether readers can reach trustworthy guidance, not just whether a file exists.

Define the project and version first. Include owned documents outside the main docs directory where appropriate, and identify generated files separately. A report can explain why a problem matters, which file supports the finding and which areas remain unchecked.

What it does

  • Follow a practical task from a README or agent entry point.
  • Distinguish a broken path from a link to a superseded version.
  • Identify competing sources that claim to govern one topic.
  • Check whether important material is reachable when it is needed.
  • Separate a confirmed defect from a question for the owner.

Install and use

  1. Use the official docs-governance plugin, shared resources and a permitted repository. Git can supply history, and optional local checks use Python 3.
  2. Ask, “Can a contributor find the current migration instructions? Report the path and evidence for each obstacle without editing files.”
  3. Review the findings and authorise any subsequent repair separately.

FAQ

Source

Made by qshanx. Original skill: docs-audit, published on GitHub under the MIT licence.

View on GitHub