principle-attack-the-premise
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides conceptual guidelines for debugging and problem-solving without any executable scripts or command-line instructions.
- [SAFE]: No network access, file system modifications, or credential handling are present in the provided files.
- [SAFE]: The configuration in
SKILL.mdexplicitly disables model invocation, and the OpenAI policy restricts implicit invocation, reflecting a secure-by-default posture.
Audit Metadata