maintain-skill-catalog

To audit and validate skill metadata across a repository, run a local auditing workflow that scans top-level skill folders to detect missing or inconsistent SKILL.md definitions and agent configurations. This process standardizes your catalog and ensures conformance.

Syncing catalog docs like README.md and AGENTS.md with actual skill capabilities requires an automated synchronization step. This workflow updates your documentation to match real skill capabilities and applies necessary validations across the entire toolkit.

Classifying and separating dependencies for installable skills involves auditing the Dependencies sections within your metadata. The validation process distinguishes between vendored, local, external, and alias dependencies to maintain catalog coherence.

Yes, you can catch missing agent configurations like openai.yaml and stale prompt references automatically. The catalog-wide validation workflow detects these issues and encodes the required fixes as durable rules within your repository.

You should run a catalog validation workflow when you need to prevent duplicates and misaligned SKILL.md definitions across your repository. It applies inventory, classification, synchronization, and validation steps to maintain overall catalog governance.

Maintaining a skill catalog does not require external dependencies. The auditing and standardization workflows operate natively on your local repository structure to validate metadata and agent configurations without third-party packages.