Make your repository legible to an agent

The prompt

Audit whether this repository is well-structured for an agent: explicit instructions, deterministic validation commands, a narrow set of tools. What's missing that would make Claude Code guess instead of know?

Why it's built that way

Claude Code is most useful when it is given a well-structured repository, explicit instructions, deterministic validation commands, and a narrow set of tools — the repository becomes the operating context and CLAUDE.md becomes the durable engineering contract.

What to do with the answer

Fix the single biggest source of ambiguity you find (usually missing or stale build instructions) before adding more agent access.