skills-discovery

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a protocol for locating and accessing skill definitions via tools like skillfind and rulesfind.\n- [SAFE]: It references local infrastructure (~/.agents, ~/.claude) and a localhost-only API (http://127.0.0.1:3401) for its operations. All paths and endpoints are consistent with the purpose of local skill management.\n- [SAFE]: It mentions an optional 'guard' hook (skills-discovery-guard.py) designed to prevent agents from inefficiently listing large directories, which is a common strategy to save tokens and maintain context relevance.\n- [SAFE]: The installation and maintenance instructions involve local script execution (install.py, build-index.py) typical for local developer tooling. No external downloads or remote code execution patterns were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 09:56 PM
Security Audit — agent-trust-hub — skills-discovery