red-team-verifier-patrick-munro
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security threats detected. The skill defines a structured process for legal document review, including factual verification and citation checking.\n- [EXTERNAL_DOWNLOADS]: The skill utilizes the agent's web search capability to access trusted legal databases and official government websites (e.g., EUR-Lex, .gov, .europa.eu) for verification purposes. This is documented neutrally as it aligns with the skill's primary function.\n- [PROMPT_INJECTION]: The skill does not contain instructions to bypass safety filters or override agent behavior. It specifically instructs the agent to adopt an adversarial mindset to detect errors and hallucinations in provided content, which acts as a safety mechanism.
Audit Metadata