The Agent Skills Directory

[SAFE]: The skill implements a routing mechanism for dispatching agent requests to specific skills based on metadata matching. The logic includes clear priority levels and tiebreaking rules (scope and type) to resolve ambiguity in a deterministic manner.
[SAFE]: Tool access is limited to 'Read' and 'Grep', which are restricted and appropriate for a skill designed to examine file system paths and skill metadata.
[SAFE]: The skill refers to the author's own repository (jacob-balslev/skill-graph) for project context and artifact verification, which is standard behavior for vendor-provided skills.

skill-router