Decision-assurance for AI agents: an auditable action boundary + receipt before it acts.
This MCP has no overview available.