Projects with this topic
Sort by:
-
Authorization and execution control for AI agents — independent policy and judge approval before consequential tool actions execute. MCP gateway, capability tokens, auditability, and fail-closed execution.
Updated
