Agent applications need specialized security incident response tools—Origin shows how to automatically map between agent-level behavior and code-level execution to quickly find attack entry points.
Origin is a security tool that helps analysts trace the source of attacks in AI agent applications by bridging the gap between high-level agent behavior logs and low-level code execution. When a security incident occurs, analysts use Origin to query trajectory logs and identify where the compromise started, without manually inspecting every line of code.