Offensive security framework for AI agent infrastructure - BloodHound for the agentic stack
Autonomous red team agent powered by LLMs for offensive security testing