grill-core
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structural and behavioral guidelines for AI agents without introducing executable code or dangerous operations.
- [SAFE]: Explicitly restricts Bash usage to read-only inspection commands (find, ls, cat, etc.), prohibiting file modifications, deletions, or network calls.
- [SAFE]: Includes a clear instruction to the agent to treat all input data as untrusted and avoid following instructions embedded in the analyzed codebases, which is a proactive measure against indirect prompt injection.
Audit Metadata