skill-issue
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The instructions direct the agent to avoid 'approval, review, or pause' and explicitly state 'Never cite it to... wait for human review'. This promotes autonomous behavior and the potential bypassing of human-in-the-loop safety checkpoints.
- [COMMAND_EXECUTION]: The documentation defines a 'Call' primitive for running scripts at specified paths ('run script to '). Although no malicious scripts are included in this file, the skill establishes a framework for external script execution without detailing verification requirements.
Audit Metadata