MCP Toolbox for Databases (BigQuery)

Add to your client

Copy the config for your MCP client and paste it into its config file.

{
  "mcpServers": {
    "mcp-toolbox-for-databases-bigquery": {
      "command": "/absolute/path/to/toolbox",
      "args": [
        "--prebuilt",
        "bigquery",
        "--stdio"
      ],
      "env": {
        "BIGQUERY_PROJECT": "<your-gcp-project-id>"
      }
    }
  }
}

Before you start

• The toolbox binary (download from GitHub releases and `chmod +x`), or run via container/source; MCP Toolbox v0.7.0 or later for the full BigQuery toolset
• A Google Cloud project with the BigQuery API enabled
• Application Default Credentials configured locally (e.g. `gcloud auth application-default login`) or a service account key
• IAM roles on the project/dataset such as BigQuery User and BigQuery Data Viewer (read), plus Job User to run queries
• An MCP-capable client (Claude Desktop, Cursor, Gemini/ADK, etc.)

About MCP Toolbox for Databases (BigQuery)

Tools & capabilities (10)

When to use it

• Use it when you want an agent to explore an unfamiliar BigQuery project by listing datasets/tables and reading schemas before writing SQL
• Use it when you need a read-only analytics assistant over your warehouse with proper IAM scoping
• Use it when you want time-series forecasting or contribution analysis driven from natural language
• Use it when you'd rather configure tools declaratively in tools.yaml than build a bespoke MCP server
• Use it when you need production features like ADC auth and OpenTelemetry tracing for an agent-to-DB bridge
• Use it when you want the same toolbox to also reach AlloyDB, Cloud SQL, Spanner, or Postgres/MySQL alongside BigQuery

Quick setup

1. 1Download the toolbox binary from GitHub releases (or use the container image) and make it executable
2. 2Enable the BigQuery API and run `gcloud auth application-default login` to set up ADC
3. 3Use the prebuilt BigQuery toolset (or write a tools.yaml with a bigquery source and the tools you want)
4. 4Start the toolbox pointing at BigQuery, then register it in your MCP client as a stdio command
5. 5Restart the client and confirm the BigQuery tools appear
6. 6Verify by asking the agent to list dataset IDs and then run a simple SELECT

MCP Toolbox for Databases (BigQuery) FAQ