mcp-compare-zipcodes
A FastMCP-compatible **Model Context Protocol (MCP)** server for comparing zipcode-level demographic distributions. It includes a tool for detecting statistically significant differences in population segments—based on age, gender, and ethnicity—between two ZIP Code Tabulation Areas (ZCTAs), using z-score analysis over synthetic census data.
Compare-Zipcodes MCP Server
A FastMCP-compatible Model Context Protocol (MCP) server for comparing zipcode-level demographic distributions. It includes a tool for detecting statistically significant differences in population segments—based on age, gender, and ethnicity—between two ZIP Code Tabulation Areas (ZCTAs), using z-score analysis over synthetic census data.
This repository also includes instructions on how to integrate the MCP server with Claude Desktop.
Features
- Compares zip code demographics based on
agexgenderxethnicityusing Z-score based anomaly detection - Synthetic population and household census data integration via
.pklfiles generated from AgentTorch
Installation
git clone https://github.com/anna8murphy/mcp-compare-zipcodes.git
cd mcp-compare-zipcodes
# install uv
curl -LsSf https://astral.sh/uv/install.sh | sh
# set up environment
uv venv
source .venv/bin/activate
uv add "mcp[cli]" httpx
Start the server
Run the server:
uv run compare-zipcodes.py
Running the server will make the compare_regions tool available, which returns a summary of statistically significant demographic differences (z-score > 2) between two ZIP codes based on age, gender, and ethnicity groupings.
Usage on Claude Desktop
Open the Claude for Desktop App configuration in a text editor:
~/Library/Application\\ Support/Claude/claude_desktop_config.json
Add the server:
{
"mcpServers": {
"mcp-compare-zipcodes": {
"command": "uv",
"args": [
"--directory",
"/ABSOLUTE/PATH/TO/PARENT/FOLDER/mcp-compare-zipcodes",
"run",
"compare-zipcodes.py"
]
}
}
}
Example Prompt: Compare the demographic statistics of the following zipcodes: 02139 and 10036.
Project Structure
mcp-compare-zipcodes/
├── README.md # documentation
├── compare-zipcodes.py # main script for Claude integration
├── server.py # MCP server implementation
├── pyproject.toml # Python project configuration
└── uv.lock # uv package manager lock file
Recommend MCP Servers 💡
datalayer/jupyter-earth-mcp-server
MCP server implementation providing geospatial analysis tools for Jupyter notebooks
lincw/dwd-mcp-server
A simple MCP server connecting Claude Desktop to the Deutsche Wetterdienst (DWD) API for German weather data
openai-server
Query OpenAI models directly from Claude using MCP protocol.
alltrails-mcp
A Model Context Protocol (MCP) server that provides access to AllTrails data, allowing you to search for hiking trails and get detailed trail information directly through Claude Desktop.
mcp-kql-server
An MCP server for executing KQL queries on Azure Data Explorer with AI schema caching and visualization.
kaeosdesign/remote-mcp-server
A remote MCP server deployed on Cloudflare Workers with OAuth support, using SSE transport for MCP client connections.