Federal Education Data MCP Server

An MCP (Model Context Protocol) server providing access to federal education data from NCES CCD, EdFacts, NCHE, and the U.S. Census Bureau. Built for AI-assisted education research and policy analysis.

Overview

Tools to search districts and schools, retrieve demographics and attendance data, access homeless-student counts, compare states on key metrics, and pull census-level demographic context. Used in production for Maestro Press research on Texas school finance and federal-funding equity.

Tools

District / school lookup

• fed_search_districts - Search LEAs by name, state, or NCES ID
• fed_search_schools - Search schools by name, state, or NCES ID

Demographics and outcomes

• fed_get_demographics - Demographic breakdown for district or school
• fed_get_chronic_absenteeism - Chronic absenteeism from EdFacts
• fed_get_homeless_data - Homeless student data from NCHE
• fed_get_census_demographics - Census Bureau ACS demographic data
• fed_get_census_district_profile - District-level census profile

Comparison and discovery

• fed_compare_states - Compare metrics across states
• fed_list_states - List US states with FIPS codes and ESC region context
• fed_get_available_years - Available school years by data source
• fed_list_available_files - Downloadable data files
• fed_download_dataset - Download and cache a specific dataset

Data Sources

• NCES Common Core of Data (CCD): https://nces.ed.gov/ccd/files.asp
• EdFacts: https://www2.ed.gov/about/inits/ed/edfacts/data-files/index.html
• NCHE (National Center for Homeless Education): https://nche.ed.gov/data/
• U.S. Census Bureau ACS: https://www.census.gov/programs-surveys/acs.html

Install

git clone https://github.com/kh0pper/fed-gov-data-mcp.git
cd fed-gov-data-mcp
uv sync

The Census Bureau API requires an API key. Set CENSUS_API_KEY in your environment or in your MCP client's env block.

Use as an MCP Server

Add to your MCP client configuration (e.g., Claude Code's ~/.claude/settings.json or a project .mcp.json):

{
  "mcpServers": {
    "fed-gov-data": {
      "command": "uv",
      "args": ["run", "--directory", "/path/to/fed-gov-data-mcp", "fed-gov-data-mcp"],
      "env": {
        "CENSUS_API_KEY": "your-census-api-key"
      }
    }
  }
}

Architecture

• FastMCP for MCP protocol handling
• SQLAlchemy (async) + aiosqlite for caching downloaded datasets
• Pydantic for response schemas
• pandas for data manipulation
• research-integration library for citation logging (sister project)

License

MIT License. See LICENSE.

Status

Released alongside ongoing research at Maestro Press. The repo has a local-path dependency on a sister library (research-integration) that is not yet published; the MCP server runs locally as configured here, but direct pip install from a clean environment will fail until that dependency is hosted publicly. Source is available for reading, forking, and reference.