authority-triage-en
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security threats were identified. The skill incorporates several security best practices: 1. Least Privilege: The configuration uses the allowed-tools field to restrict the agent to Read access only, preventing unauthorized file system modifications or network exfiltration. 2. Human-in-the-loop: The workflow explicitly mandates a 'Human gate' where a qualified professional must review and approve the generated source map before it is used. 3. Data Integrity: The instructions require that any article number not confirmed in a database must be tagged with [TO VERIFY], reducing the risk of hallucinations or reliance on unvalidated information.
Audit Metadata