skill-registry

To index available AI skills, you scan project and user skill directories to discover SKILL.md files, extract frontmatter metadata, deduplicate entries, and generate a searchable skill registry output file.

Keeping track of moved or renamed AI skills requires refreshing your skill registry to scan directories, update preserved skill paths, and ensure the indexed metadata accurately reflects the new locations.

Skill discovery works by scanning specified project and user directories for SKILL.md files, extracting their frontmatter metadata, and using it to build an indexed registry of available AI agents.

You need to regenerate the skill registry whenever you install, remove, create, or move AI skills, ensuring the indexed metadata stays synchronized with the actual SKILL.md files across your directories.

Yes, the skill registry uses SKILL.md as the source of truth, extracting frontmatter metadata from these files to build the index while preserving the original skill paths for accurate discovery.

Yes, you can deduplicate skill entries by scanning both project and user skill directories, extracting metadata, and filtering redundant records to maintain a clean and accurate registry output file.