A
license-
qualityC
maintenanceProvides a human-in-the-loop security layer for AI agents by intercepting file operations, explaining them with a local LLM, and enforcing a deterministic policy that requires user approval for risky actions.
Last updated
MIT