frame
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill instructions are focused on a manual auditing process without automated execution of external code or access to restricted system resources.- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided declarations and existing project documentation (PRODUCT.md) for auditing. The risk of indirect injection is mitigated by explicit instructions to never invent users, outcomes, or evidence, and to only report findings based on what the builder has explicitly stated.
Audit Metadata