sani903/OpenAgentSafety

sani903/OpenAgentSafety provides a ready‑made framework for testing AI agent safety in realistic environments, letting teams add robust safety evaluation withou