Organization Policy Adds MCP Server and Gemini Custom Constraint Authoring

Google Cloud‘s Organization Policy Service addresses a common root cause of security incidents: misconfigured cloud resources. It lets administrators set security guardrails across the resource hierarchy, drawing on a catalog of more than 100 predefined policies and on Custom Organization Policy for granular control over resources.

To make guardrails easier to author, audit, and scale, Google Cloud introduced the general availability of an Organization Policy Model Context Protocol (MCP) server and natural-language constraint authoring powered by Gemini. The MCP server lets autonomous agents and pipelines interact directly with Organization Policy instead of manually parsing hierarchy chains. Its tools cover constraint discovery (ListConstraints, GetCustomConstraint), policy evaluation (GetEffectivePolicy calculates the effective policy after hierarchy inheritance and tag-based conditions), and policy management (CreatePolicy, CreateCustomConstraint). Agents can use these capabilities for posture audits, drift detection, and exemption management, and administrators can also build custom agents and skills to streamline policy management. A described workflow combines a preferred coding agent with the Organization Policy MCP tools to create a custom dashboard that visualizes enforced constraints and identifies folders and projects exempted from a policy.

Custom Organization Policy previously required writing rules in Common Expression Language (CEL) with a detailed understanding of Google Cloud resource schemas. With the new Gemini integration, administrators can author custom constraints in natural language and refine them conversationally. Examples include requiring agent-identity in Agent Runtime, requiring that agents deployed to Agent Runtime use an approved Agent Gateway, and preventing users from disabling auto-upgrades on Google Kubernetes Engine. To support safe deployments, Gemini can generate compliant and non-compliant test resource configurations for validating policy in dry-run mode before enforcement. Google Cloud Professional Services also provides a sample custom constraints library with pre-authored samples based on CIS and PCI standards. Custom constraints are supported on a range of services, including Agent Runtime, Agent Search, and Agent Studio on the Gemini platform, Cloud Run, and GKE.

The broader message is that Organization Policy offers proactive governance of resource configurations: administrators enforce critical guardrails so engineering teams can build quickly without bypassing security. The combination of an expanding built-in guardrail catalog, autonomous tools, and custom constraints across Google Cloud services is meant to let enterprises scale safely.

AI-powered cloud governance: What's new in Organization Policy

View Original