skill-registry

To maintain a centralized skill registry, you can automate scanning user-level and project-level skill definitions to build a comprehensive catalog. This ensures consistent discovery and execution by the orchestrator across your workspace.

Deduplication logic prioritizes project-level skill definitions when conflicts arise during the scan. This ensures that project-specific configurations reliably override user-level defaults in the final registry.

Automating skill discovery involves scanning your environment for skill definitions and persisting the results. The orchestrator can then reference the generated registry file to reliably find and execute available skills.

The automated skill registry saves its data by writing to the .atl/skill-registry.md file. When available, it also persists the registry to engram storage for cross-session usage.

Yes, the skill registry specifically applies to multi-repo environments where both user-level and project-level skills exist. It covers installation, updates, and de-duplication workflows across these repositories.

Your orchestrator may fail to find new skills if the registry is out of date. Running an automated scan updates the .atl/skill-registry.md file, ensuring the orchestrator can reference all available skills reliably.