The tool reasons like a human researcher, traces data flows, and assigns severity ratings after multi-stage verification. It emphasizes human-in-the-loop approval with confidence scores. Aimed at countering AI-assisted attacks, it launched February 21, 2026.