pentest-input-protocol-manipulation
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill file contains high-level instructional content defining how an agent should approach security testing. It does not include scripts, commands, or tools that execute code or perform network activities.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for testing untrusted inputs, which is an inherent part of penetration testing tasks. However, the instructions are purely methodology-based and do not provide automated exploit code or specific vulnerable interpolation patterns. The risk is minimized by the instructional nature of the content.
Audit Metadata