codex-skill-audit

To audit a Codex skill repository for structure and frontmatter quality, run the deterministic scripts to scan the .agents/skills tree for missing frontmatter, kebab-case violations, and long SKILL.md bodies, generating both human-readable and machine-readable summaries.

Repository structure health for Codex skills ensures governance readiness by verifying that frontmatter, kebab-case naming, and operator sections meet standards. Maintaining these standards prevents structural gaps and ensures reproducible evaluation across the entire skills tree.

Yes, you can validate Review Harness configuration automatically. The audit verifies Review Harness sections, axes, and actions across the repository to ensure governance and reproducible evaluation are correctly configured.

You can check for missing frontmatter and kebab-case violations by running the deterministic audit scripts against the target repository. The scripts detect these structural gaps across all skills under the .agents/skills directory and report them in the findings.

Yes, the audit generates a machine-readable summary for automated dashboards. It aggregates evidence-driven findings, strengths, and structural gaps into outputs suitable for automated pipelines and governance reporting.

Deterministic repository audits for skill governance readiness are limited to detecting structural gaps, frontmatter quality, and Review Harness configuration. They do not evaluate the semantic logic or runtime behavior of the skills within the .agents/skills tree.