Empower your agents with the Google Cloud CLI remote MCP server

Google Cloud unveiled a preview of a managed remote MCP server that hosts the gcloud and bq command‑line tools. The service runs inside an isolated sandbox on Google Cloud infrastructure, letting AI agents invoke any of the hundreds of CLI commands without installing binaries locally. By implementing the Model Context Protocol, the server can be reached from any MCP‑compatible platform through standard configuration. The remote server removes the need for agents to maintain local CLI versions, which reduces operational overhead for teams building custom agents. Because the execution environment is network‑restricted and uses zero ambient credentials, security and governance are enforced through IAM, OAuth 2.0 and Model Armor. Auditing is built in, with every invocation recorded in Cloud audit logs, giving administrators full visibility into caller identities and authorization decisions. It is not yet certain how performance will compare with locally installed CLIs, and Google has not published quantitative benchmarks. The preview status means that availability, pricing and long‑term support are still undefined, and the extent of compatibility with all existing gcloud and bq commands remains to be proven.