implementation-skill-routing
Pass
Audited by Gen Agent Trust Hub on Aug 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates as a selector for other skills within a project-specific registry. It includes explicit constraints to prevent command injection or path traversal by only returning validated skill names and forbidding the return of file paths or direct loading of skill bodies.
- [INDIRECT_PROMPT_INJECTION]: The skill reads external data from
.ai/atl/skill-registry.md. However, it implements defensive measures including strict column matching, a limit of three results, and a prohibition on inventing names or using location data for execution. This significantly reduces the risk of data-driven behavior overrides. - [COMMAND_EXECUTION]: No shell commands or direct execution patterns were detected. The skill acts purely as a logic layer for plan generation.
Audit Metadata