set-threshold
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides instructions for a decision-making process. No evidence of prompt injection, obfuscation, or unauthorized data access was found.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines an ingestion surface for external data (criteria, evidence, decision context), but poses no risk as it lacks exploitable capabilities.\n
- Ingestion points: Processed as inputs defined in the
SKILL.mdinput contract.\n - Boundary markers: None present.\n
- Capability inventory: None. The skill does not use shell commands, network tools, or file-writing operations.\n
- Sanitization: None present.
Audit Metadata