agent-evaluation-harness

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill contains only documentation and behavioral guidelines for AI agents to follow when assisting with agent evaluation. No malicious instructions, sensitive data access, or network calls were detected.
  • [NO_CODE]: The skill consists entirely of Markdown files providing instruction and examples. There are no scripts, binaries, or automated installation steps included in the analyzed content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:33 PM
Security Audit — agent-trust-hub — agent-evaluation-harness