HandoffProbe
Open-source security testing for AI agent handoffs
Visit Website
AI agent systems can look secure in isolation and still lose security guarantees at the handoff. I built HandoffProbe to make those boundary failures reproducible and testable.
