Connects Claude to Google BigQuery through the Model Context Protocol. Exposes three core operations: listing tables, fetching schemas, and executing arbitrary SQL queries. Resources provide browsable access to table metadata at `bigquery://tables` URIs, while the query tool lets Claude run SQL directly against your datasets. Built by SnowLeopard AI as a benchmarking tool for their platform, now open sourced. Authentication uses Google Cloud CLI credentials, and you configure it by pointing at specific datasets or tables in your project. Supports both public datasets like `bigquery-public-data.usa_names` and your private tables. Reach for this when you need to give Claude read access to your data warehouse without building custom extraction pipelines.
claude mcp add --transport stdio snowleopard-ai-bigquery-mcp uvx bigquery-mcp