What problem does it solve?
This Skill helps you detect mismatches between a repo’s operational documentation (CLAUDE.md, AI_INDEX.md, AGENTS.md, and local skill docs) and what the project actually implements, preventing incorrect agent behavior.
Core Features & Use Cases
- Cross-file coherence checks: Compares CLAUDE.md, AI_INDEX.md, AGENTS.md, and local
.agents/skills/*/SKILL.md against the repository’s current codebase and structure.
- Scope-limited auditing: Focuses specifically on documentary consistency rather than deep code or per-skill technical correctness.
- Truth-first reporting: Produces only inconsistencies verified against real files, minimizing false assumptions.
Use Case: When onboarding or refactoring an automation-heavy repo, run this audit to ensure the “operational map” used by AI agents still matches reality and to avoid prompts that could cause incorrect or unsafe workflows.
Quick Start
Ask the auditor: "Audit document consistency between CLAUDE.md, AI_INDEX.md, AGENTS.md, and local skills, and report only real inconsistencies with severity and impact."