ca.structureclerk/mcp

AI agent policy decisions: ALLOW, DENY, APPROVE, ESCALATE.