debate
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides logic for managing inter-agent communication during a debate. It does not exhibit malicious behaviors or attempt to access sensitive resources.\n- [PROMPT_INJECTION]: The instructions are designed to frame debates and moderate content. There are no attempts to bypass safety filters or ignore system instructions.\n- [DATA_EXFILTRATION]: No data exfiltration vectors, such as network calls to external domains or sensitive file reading, were found.\n- [REMOTE_CODE_EXECUTION]: The skill does not reference or execute code from external or untrusted sources.
Audit Metadata