Hostile-by-default critic that grades AI agent output against your acceptance policy.
This MCP has no overview available.