army-agent-contract
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of documentation and guidelines for defining agent boundaries. It emphasizes isolation via disjoint write allowlists and explicitly forbids agents from modifying their own security policies or evaluation criteria. No executable code, network calls, or risky system operations were found.
Audit Metadata