git-skill-manager

Git-based skill lifecycle management uses repository registration, pull, and push workflows to keep agent skills current across teams. It relies on SKILL.md frontmatter for metadata, a central registry for tracking, and scripts plus references directories to execute cross-repo discovery, evaluation, and auto-update operations.

Cross-repo skill discovery and merging is handled through Git-based lifecycle operations that track feedback, evaluate, and refine skills across multiple repositories. The process uses a central registry and auto-update workflows to merge and synchronize skill versions, keeping them fresh across distributed team environments.

To register a Git repository and pull a skill, you add the repository to the skill registry and execute a pull command to retrieve the skill files. The pulled skill is then managed using its SKILL.md frontmatter for identification, along with its scripts and references components for operational execution.

Yes, Git automation supports pushing updated skills to a registered team repository. The lifecycle manager executes push commands that send refined or evaluated skill updates back to the team repository, ensuring all members access the latest skill versions through synchronized registry workflows.

No external dependencies are required for the skill lifecycle manager to perform Git operations. It operates independently using built-in scripts and references components to handle repository registration, pull, push, and auto-update workflows directly through standard Git commands.

Git-based lifecycle management for agent skills is not suitable for scenarios requiring real-time skill updates or centralized non-versioned storage. It is designed for cross-repo workflows using SKILL.md frontmatter and Git repositories, making it less ideal for environments not already using Git for version control.