capability-recommender
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified. The skill acts as a documentation-based assistant and does not execute system commands or access sensitive information.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided task descriptions to identify capability gaps, which could contain untrusted instructions. 1. Ingestion points: User-provided task/workflow gap descriptions ingested via SKILL.md. 2. Boundary markers: The skill body contains explicit instructions to perform a security audit on all candidates, noting 'red flags' for behavior. 3. Capability inventory: No write, network, or execution capabilities are defined or requested in the skill metadata or body. 4. Sanitization: The agent is directed to check for patterns like unauthorized network access or credential reading in recommended tools before approval.
Audit Metadata