onboarding

This skill checks AGENTS.md instructions against repository-specific best practices and source-of-truth docs to identify gaps in operating rules. It requires an AGENTS.md file plus canonical docs, then summarizes missing or stale guidance before applying approved updates.

Mapping source-of-truth documents requires pointing agents to canonical docs that govern delivery workflow, architecture, and validation. This skill evaluates AGENTS.md against those repository docs to ensure the operating rules and source-of-truth map remain clear and compact.

You cannot update repository instructions without explicit approval. This skill recommends or updates AGENTS.md guidance only after receiving approval, ensuring changes remain minimal, retrieval-oriented, and controlled before broader work begins.

You need an AGENTS.md instruction critique when a repository requires first-pass agent onboarding or a controlled cleanup of operating rules. This applies before deeper work begins, ensuring the source-of-truth map and guidance are clear and compact.

Identifying stale or missing guidance requires checking AGENTS.md operating rules against repo-specific best practices and canonical docs. This skill compares the instructions to source-of-truth documents to find gaps, requiring the AGENTS.md file and relevant docs for analysis.