enhance-skill
Pass
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The workflow is transparent and includes human-in-the-loop validation for all filesystem modifications.
- [PROMPT_INJECTION]: The skill processes external data (other skill files), creating a surface for indirect prompt injection. This is a characteristic of its intended function as an enhancement tool. The risk is managed via user confirmation before any modifications are applied.
- Ingestion points: Reads SKILL.md and supporting files in Step 1.
- Boundary markers: Not specified.
- Capability inventory: Read and write access to files within the user-specified skill path.
- Sanitization: The skill focuses on structural and clarity improvements without explicit content sanitization.
Audit Metadata