mcp-census
@aaronbrezel
Proof of concept for an MCP Server delivering Census Bureau data for AI Agent interoperability
Overview
What is mcp-census?
mcp-census is a proof of concept for an MCP (Model Context Protocol) server that delivers Census Bureau 2020 decennial data for AI Agent interoperability. It is intended for developers building AI agents that need to access U.S. census data via the MCP protocol.
How to use mcp-census?
The server requires two API keys: CENSUS_API_KEY (request at https://api.census.gov/data/key_signup.html) and GEMINI_API_KEY (request at https://ai.google.dev/gemini-api/docs/api-key). To run the MCP server, execute uv run python mcp_server/app.py. An accompanying agent can be run with uv run python app.py.
Key features of mcp-census
- Delivers 2020 decennial Census Bureau data
- Integrates with Gemini API for AI agent capabilities
- Uses the Model Context Protocol (MCP)
- Includes a sample agent for proof-of-concept use
- Operates with the
uvPython package manager
Use cases of mcp-census
- AI agents querying U.S. 2020 census data via natural language
- Prototyping MCP-based interoperability between census data and AI models
- Demonstrating how to connect government statistical data to conversational agents
FAQ from mcp-census
What API keys are required?
You need a Census API key and a Gemini API key. Sign up at the links provided in the README.
How do I start the MCP server?
Run uv run python mcp_server/app.py in your terminal.
Is mcp-census ready for production?
No, it is a proof of concept, not a production-grade service.
What data does the server serve?
It serves Census Bureau 2020 decennial data. No other datasets are mentioned.
What runtime does mcp-census require?
It uses Python and the uv package manager for execution.
