BigQuery MCP Server

This is a Model Context Protocol (MCP) server designed to provide Large Language Models (LLMs) with access to BigQuery. It allows LLMs to inspect database schemas and execute queries against BigQuery datasets.

Components

Tools

The server implements the following tools:

• execute-query: Executes SQL queries using the BigQuery dialect.
• list-tables: Lists all tables within the BigQuery database.
• describe-table: Describes the schema of a specified table.

Configuration

The server requires configuration with the following arguments:

• --project: The GCP project ID.
• --location: The GCP location (e.g., europe-west9).
• --dataset (optional): Specifies BigQuery datasets to consider. Multiple datasets can be specified by repeating the argument. If omitted, all datasets in the project are considered.

Quickstart

Install

Installation instructions are provided for both Smithery and manual configuration with Claude Desktop.

Development

Instructions are provided for building, publishing, and debugging the server, including the use of uv for dependency management and the MCP Inspector for debugging over stdio.