Google Cloud CLI remote MCP server
A Google-managed remote MCP server, in preview, that runs Google Cloud CLI commands for an AI application in a sandbox on Google's side.
What it does
A Google-managed remote MCP server, in preview, that runs Google Cloud CLI commands for an AI application in a sandbox on Google's side. It is turned on by enabling the Cloud CLI Execution API and is reached at cloudcli.googleapis.com/mcp. It offers a tool for gcloud commands and one for bq commands, so an agent can provision resources, read logs and diagnose problems, or manage BigQuery datasets, jobs, schedules, reservations and access. Some command groups such as auth, billing, config and components are refused. Sign-in uses OAuth and IAM, and Google's guide covers clients including Antigravity, ChatGPT and Claude.
Capabilities
- Run gcloud commands against a Google Cloud project
- Run bq commands for BigQuery datasets and jobs
- Create, update or delete Google Cloud resources
- Pass input files to commands that read configuration
Details
| Fact | Value | Source |
|---|---|---|
| Type | MCP server | |
| Tool | Google Cloud | |
| Publisher | Google Cloud | source |
| Official status | Officialdocumented on Google Cloud's own docs site | source |
| Status | Preview | source |
| Connection | Remote (hosted, you connect over the internet) | source |
| Can change data | Yes | source |
| Needs an API key or sign-in | Yes | source |
| Platforms | Not stated in its documentation | |
| AI clients | Not stated in its documentation | |
| Licence | Not stated in its documentation | |
| Pricing | Not stated in its documentation | |
| Activity | No commit dataDownload data not available (not on npm or PyPI) | |
| Links | Website |