agent-guidelines

You can generate AI coding guidelines by scanning your repository's technology stack to produce structured, path-scoped rules and reference docs. It uses discovery scanners for build, test, and source patterns to deliver actionable coding rules.

Path-scoped rule generation creates structured guidelines targeting specific directories within multi-language repositories. It scans your tech stack and source patterns so AI agents apply the correct framework conventions to the right project paths.

Yes, you can audit and update existing coding rules during framework migrations. The Skill includes a rules auditor scanner that evaluates current guidelines against discovered configurations, ensuring AI reference docs remain accurate.

Yes, this tool supports multi-language repositories adopting new frameworks or undergoing tool migrations. It combines discovery scanners to identify diverse tech stacks and generate corresponding guidelines for each detected language.

The best way to maintain CLAUDE.md pointers is to update them automatically during guideline generation. This Skill updates CLAUDE.md to reference newly generated Layer 1 rules and optional Layer 2 docs, keeping AI agents aligned.

You should avoid automated reference doc generation when a repository lacks stable build or test configurations, as discovery scanners require identifiable patterns. It is also less effective for single-language projects without active migrations.