evolve-skill
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses standard Unix utilities including cat, head, grep, mkdir, cp, ln, find, wc, readlink, and git to manage skill files, update version metadata, and maintain a local registry.
- [SAFE]: No network operations or external downloads (curl, wget, etc.) were identified; all operations are performed on the local filesystem.
- [SAFE]: The skill manages symbolic links within the project directory and the user's local Claude configuration directory (~/.claude/skills), which is consistent with the intended purpose of managing local agent skills.
- [SAFE]: No hardcoded credentials, sensitive file access (e.g., SSH keys, AWS config), or obfuscation techniques were found.
Audit Metadata