skill-manager

To check if your GitHub-wrapped skills are outdated, the tool scans your local directory for SKILL.md metadata, fetches remote commit hashes via git ls-remote, and compares them against local github_hash values to classify entries as current, outdated, or error.

Yes, you can batch update outdated GitHub skills at once using a structured workflow with guided prompts, upgrading selected entries after scanning your local library and classifying which ones require updates.

For automated git updates, a skill needs the github-to-skills metadata standard, requiring both github_url and github_hash fields specified inside its SKILL.md file to enable remote commit hash comparison.

To manage and clean up a large local library of skills, list all local entries using inventory scripts, identify outdated or error states through automated scanning, and delete unwanted skills through a structured cleanup workflow.

Yes, the skill audit process generates JSON-ready reports for automation pipelines, outputting structured data after evaluating local SKILL.md metadata against remote commit hashes for large skill collections.

Yes, you need pyyaml installed to scan local skills for metadata, as the dependency is required to parse SKILL.md files and extract github_url and github_hash fields during the automated audit process.