BigQuery Biomedical MCP

Example Use Cases

• 1

  Execute controlled SELECT queries against public biomedical BigQuery datasets while respecting dataset and cost restrictions.

• 2

  Explore dataset structure and table schemas (column names, types, nullability) to plan accurate SQL queries.

• 3

  Record, append, and retrieve session-level analysis insights to maintain contextual state during multi-step analyses.

Description

This project implements a Model Context Protocol server tailored to biomedical BigQuery datasets, exposing tools to list datasets, run SELECT queries, inspect table schemas, and record analysis findings. It includes a database resource (OpenTargets schema), a memo resource for session insights, and environment-configured controls like allowed datasets and query cost limits. Intended for biopharma-focused LLM agents, it helps plan and execute queries while preventing runaway BigQuery costs. The code is open-source (GPL-3.0) and requires a Google Cloud service account with BigQuery access.