politics-coach
Pass
Audited by Gen Agent Trust Hub on Feb 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): No security issues detected. The skill operates as intended for stakeholder simulation and organizational coaching.
- [DATA_EXFILTRATION] (SAFE): The skill has no network capabilities and does not attempt to send data to external domains.
- [COMMAND_EXECUTION] (SAFE): No shell commands or system-level scripts are executed; activities are restricted to reading and writing markdown documents.
- [PROMPT_INJECTION] (SAFE): The instructions do not contain patterns for bypassing agent safety filters, ignoring previous instructions, or extracting system prompts.
- [INDIRECT_PROMPT_INJECTION] (SAFE): While the skill ingests data from stakeholder avatar files (01-Company-Context/1.1-Stakeholder-Avatars/), this is used solely for persona simulation. The skill lacks capabilities (like network or command execution) that could be exploited via malicious data in those files.
Audit Metadata