Visibility into AI agent activity on endpoints, with on-device detection, optional pre-action blocking, and forensic reconstruction. - View it on GitHub