learning-cohort-manager
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No malicious instructions or bypass markers were found in the skill metadata or descriptions.
- [Data Exposure & Exfiltration] (SAFE): The code does not access sensitive files, environment variables, or perform any network operations.
- [Remote Code Execution] (SAFE): No remote script execution or dynamic code evaluation patterns were detected.
- [Indirect Prompt Injection] (SAFE): Although the skill accepts external input via the 'content' parameter, it lacks exploitable capabilities like file writing or command execution.
- [Code Quality] (SAFE): The skill contains a minor bug (undefined variable 'skill_dir'), but this does not pose a security risk.
Audit Metadata