pm-debate

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security risks were identified in the skill definition.
  • [PROMPT_INJECTION]: The instructions establish a professional persona without attempting to bypass safety filters or override system instructions. Natural language triggers like '辩论' (debate) are within the context of the product manager role-play.
  • [NO_CODE]: The skill consists entirely of natural language instructions; no scripts, binaries, or automated command executions are present.
  • [DATA_EXFILTRATION]: No commands or instructions attempt to access local files, environment variables, or perform network requests to external domains.
  • [EXTERNAL_DOWNLOADS]: The skill does not reference or download any external resources or dependencies.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 01:50 PM