An enforcement layer that validates AI agent actions against governance policies, including path permissions and content scanning, at runtime. It enables secure, role-based execution of file operations and commands with zero token overhead by processing policies independently from the agent's contex
View & run on Rokha →