config-doctor-litelisted

# Config Doctor Lite Health-check your AI coding agent's configuration. Find broken paths in CLAUDE.md, skill folders the agent cannot load, and frontmatter that hurts trigger accuracy — before they silently corrupt your agent's behavior. ## When to use / when not to use **Use this skill when:** - The user asks to audit, health-check, or clean up their Claude Code configuration (CLAUDE.md, skills). - A project folder was **moved or renamed** and references in CLAUDE.md may now point at nothing. - A skill **silently stopped working** — it no longer loads or triggers, and nobody changed the skill itself. - The user migrated to a new machine and wants to verify the config survived. **Do NOT use this skill for:** - Auditing code dependencies (npm audit, pip-audit, cargo audit). - Scanning for secrets or hardcoded credentials (different product). - Linting application configs: nginx, Docker/docker-compose, ESLint, Kubernetes, CI pipelines. - Writing a new CLAUDE.md from scratch (this tool checks existing config; it does not author it). - Checking whether web URLs are alive (only local file paths are verified). ## Quick start Run the scanner first, then talk about the results: ``` python scripts/config_doctor_lite.py ``` The script is **100% read-only**. It never creates, modifies, or deletes anything, so it is always safe to run. Use `--root PATH` (repeatable) for config kept in non-standard locations. ## What gets checked (3 of the full 8) | ID | Check | Sev