safety-guard
Pass
Audited by Gen Agent Trust Hub on Apr 1, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a defensive tool designed to restrict potentially harmful agent behaviors. No malicious patterns or security risks were identified.\n- [COMMAND_EXECUTION]: The skill identifies high-risk commands (e.g.,
rm -rf,git push --force,chmod 777,sudo rm) specifically for the purpose of monitoring and preventing their execution. This is a safety feature and does not constitute an attempt to execute malicious commands.\n- [DATA_EXFILTRATION]: No network operations or external data transmissions were detected. The skill mentions logging blocked actions to a local file (~/.claude/safety-guard.log) for user auditing and oversight.
Audit Metadata