bigquery-mcp-server

bigquery-mcp-server

3.3

If you are the rightful owner of bigquery-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

A Model Context Protocol (MCP) server for accessing Google BigQuery. This server enables Large Language Models (LLMs) to understand BigQuery dataset structures and execute SQL queries.

The BigQuery MCP Server is designed to facilitate the interaction between Large Language Models (LLMs) and Google BigQuery. It provides a structured way for LLMs to comprehend dataset structures and execute SQL queries in a secure and efficient manner. The server supports authentication through Application Default Credentials (ADC) or service account key files, ensuring secure access to BigQuery resources. It offers a range of tools for executing read-only queries, listing datasets and tables, retrieving table information, and performing dry runs to estimate query costs. The server is equipped with security features to restrict operations to read-only access, preventing any modifications to the data. It also includes recommendations for partition filters to optimize query performance. The server can be installed locally or run in a Docker container, and it integrates seamlessly with MCP-enabled LLMs through a straightforward configuration process.

Features

  • Authentication and Connection Management
  • Read-only SQL Query Execution
  • Dataset and Table Listing
  • Table Information Retrieval
  • Dry Run Query for Cost Estimation

Tools

  1. query

    Execute read-only (SELECT) BigQuery SQL queries with configurable maximum results and bytes billed.

  2. list_all_datasets

    List all datasets in the project, returning an array of dataset IDs.

  3. list_all_tables_with_dataset

    List all tables in a specific dataset with their schemas, requiring a datasetId parameter.

  4. get_table_information

    Get table schema and sample data, supporting partitioned tables with partition filters.

  5. dry_run_query

    Check query validity and estimate cost without execution, returning processing size and estimated cost.