tasting
Pass
Audited by Gen Agent Trust Hub on Jun 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill acts as a set of instructions for the agent to orchestrate self-review using other available internal tools (shower, ssotchk, re0, ssotize).
- [PROMPT_INJECTION]: No evidence of instructions attempting to bypass safety protocols, override system behavior, or elicit sensitive information.
- [DATA_EXFILTRATION]: No sensitive file access or outbound network communication patterns were detected.
- [COMMAND_EXECUTION]: The skill does not perform any direct shell command execution, subprocess spawning, or system-level operations.
Audit Metadata