liffy
Fail
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFECOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill instructions involve cloning an external repository (https://github.com/mzfr/liffy) and executing its content using python3 liffy.py. This allows the execution of arbitrary code from an unverified third-party source inside the agent's execution environment.
- [EXTERNAL_DOWNLOADS]: The skill fetches tools at runtime from a GitHub repository not associated with a trusted organization, creating a supply chain risk.
- [CREDENTIALS_UNSAFE]: The skill explicitly references paths to sensitive data including /home//.ssh/id_rsa (SSH keys), /proc/self/environ (potential API keys/secrets), and /var/lib/php/sessions/ (session tokens).
- [COMMAND_EXECUTION]: The skill relies on shell command execution (git clone, python3) for downloading and running exploitation scripts.
Recommendations
- AI detected serious security threats
Audit Metadata