The Data Commons Model Context Protocol (MCP) Server

Benefits of MCP Server

The MCP Server provides a standardized way for AI agents to consume Data Commons natively. This allows developers to leverage our comprehensive data without needing to learn or directly interact with complex underlying APIs. It dramatically accelerates the creation of data-rich, agentic applications that reduce the rate of hallucinations in LLMs.

Faster than ever, developers can deploy AI agents and applications that deliver trustable, sourced Data Commons information back to the end user. The MCP Server enables agents to handle the full range of data-driven queries, from initial discovery to generative reports:

• Exploratory: "What health data do you have for Africa?"
• Analytical: "Compare the life expectancy, economic inequality, and GDP growth for BRICS nations."
• Generative: "Generate a concise report on income vs diabetes in US counties."

Ready to try it? Get started with Gemini CLI here.

Data Commons MCP Server is designed for seamless integration into agent development workflows. Here, prompted by a single query in the Gemini CLI client, an AI agent systematically fetches information across many of Data Commons’ complex datasets.

Real-World Use Case: The ONE Data Agent

Since 2023, Google's Data Commons has partnered with the ONE Campaign, a global organization that advocates for the investments needed to create economic opportunities and healthier lives in Africa. This collaboration led to the creation of ONE Data, a platform that combines ONE's global development data and policy expertise with the extensive public datasets available through Data Commons.

As our first use case, ONE Data leveraged the power of our MCP server and agent-driven exploration to develop The One Data Agent, an interactive platform for health financing data. This new tool enables users to quickly search through tens of millions of health financing data points in seconds, using plain language. They can then visualize that data and download clean datasets, saving time while helping to improve advocacy, reporting and policy-making.

This is a critical innovation for those working on global health. There is an urgent need to strengthen health systems in developing countries, but finding reliable data on health financing is a significant challenge – truly searching for the proverbial needle in a haystack. The information is scattered across thousands of disparate silos, buried in different reporting formats, organized by technical jargon and stored in several isolated databases. Now, for example, if you want to identify which countries are at risk from donor cuts, you can quickly search for countries that rely most on external funding for health and are therefore most vulnerable to aid reductions or debt shocks.

To compile a reliable report from traditional databases, users would need to work across datasets and manually pull data. Agents, however, understand complex queries and are able to fetch and compile the needed data quickly. The ONE Data Agent is paving the way for a new era of accessible, impactful data-driven advocacy.

Getting Started

Whether you’re prototyping a new AI agent, adding data features to your product or streamlining your organization’s analytics workflow, the Data Commons MCP Server is ready to help you move faster.

It’s designed for seamless integration and minimal onboarding friction. The Data Commons MCP Server fits naturally within Google Cloud Platform's latest agent development workflows, such as the Agent Development Kit (ADK) and clients including Gemini CLI. The server can also be easily integrated with any other agentic workflow or platform.

This example shows an AI agent, built using the Data Commons MCP Server, turning a simple user query into a data overview report.

To help you get started, we provide an ADK sample agent in a Colab notebook and instructions for using the Server with Gemini CLI:

• Try it out in Gemini CLI or your favorite MCP client by installing the PyPi package.
• Get started developing an ADK agent with Google Colab
• Explore our GitHub Repository to see the sample agent and start building your own.