AgentTriage Provides Replayable Security Diagnosis for Agent Runtimes
October 23, 2026
AgentTriage implements an executable contract to enable replayable security diagnosis across heterogeneous agent runtimes. It uses runtime overlays to normalize observations and constructs a typed provenance graph, allowing security claims to be re-checked without relying on LLM judgment.
HOW THIS AFFECTS YOU
●
builderYou can implement deterministic security auditing for agents that execute code or interact with files.