A2M LogoA2M
← All listings

openuspublicdata-mcp

MIT License

Federal-first, no-key MCP access to high-value US public data sources

Key Value Proposition

Gives developers and AI agents authenticated, no-key MCP access to 16 federal and state public data tools plus five curated state portals in a single stdio or HTTP server

Summary

OpenUSPublicDataMCP is a FastMCP server that exposes high-value US public data through the Model Context Protocol, covering federal APIs like Data.gov, Census, Federal Register, USAspending, and Congress.gov as well as reusable Socrata, ArcGIS, and CKAN adapters. Built for developers and AI agents who need programmatic access to government data without keys where practical, it bundles 16 MCP tools with provenance metadata, curated state portals for California, New York, Texas, Washington, and Florida, and a React/Leaflet discovery explorer. Set it up with a virtual environment and pip install, then call tools directly via stdio or HTTP transport, e.g. 'fastmcp call ... search_data_gov query=climate'.

Proof Points

  • 16 MCP tools covering federal APIs (Data.gov, Census ACS, Federal Register, USAspending, Congress.gov) plus reusable Socrata, ArcGIS REST, and CKAN adapters
  • No-key-first design: most core tools work without credentials while optional keys remain fully documented
  • Includes a React/Leaflet web discovery explorer mapping all 50 states plus DC with provenance, jurisdiction, auth, platform, licence, and coverage metadata

About

Federal-first MCP server for high-value US public data APIs

User Reviews

Genuine feedback from the developer community.

You must be signed in to leave a review.

No reviews yet. Be the first to review this project!

Are you the author of this project?

Claim this pre-seeded listing to manage details, edit tags, or upload assets.

Please sign in using the button in the header to claim repository ownership.

Share Project

Project Facts

LicenseMIT
Trust score88/100

Author & Community

AuthorStephen Phillips

List your project with one command.