title	emoji	colorFrom	colorTo	sdk	sdk_version	app_file	pinned
MCP Postal Geocoder	📍	blue	green	streamlit	1.39.0	streamlit_app.py	false

MCP Postal Geocoder

Complete MCP Server & Client Implementation for US Postal Code Geocoding

A full-featured Model Context Protocol implementation including both MCP Server (FastMCP backend) and MCP Client (Streamlit demo app). Provides a drop-in replacement for GeoNames postal code API using US Census Bureau ZIP Code Tabulation Areas (ZCTAs). Fast, accurate, and comprehensive with 33,791 US postal codes.

🌟 Features

• Complete MCP Implementation: Both server (5 tools) and client (Streamlit app)
• 5 MCP Tools: Search, geocode, reverse geocode, validate, and statistics
• Census Accuracy: Official US Census Bureau ZCTA data
• High Performance: <1ms exact lookups, <50ms reverse geocoding
• Complete Coverage: All 33,791 US postal codes with cities and states
• Reference Implementation: Perfect example for building MCP clients
• Multi-Environment: Works in development, containers, and cloud deployments

🚀 Demo

Try the live demo: MCP Postal Geocoder on HuggingFace Spaces

Features in Demo:

• 📊 Database Statistics - Real-time stats on 33,791 postal codes
• 🔍 Search Tools - Exact match and prefix search with maps
• 🌍 Geocoding - Convert postal codes to coordinates
• 📍 Reverse Geocoding - Find postal codes near any location
• ✅ Validation - Quick postal code validation

🛠️ Available Tools

1. postal_code_search - Search postal codes (exact match or prefix)
2. geocode_postal - Convert postal code to coordinates
3. reverse_geocode - Find postal codes near coordinates
4. validate_postal - Check if postal code exists
5. postal_stats - Database statistics and health

📦 Installation

pip install -e .

🎯 Usage

MCP Server

# Start MCP server
mcp-postal-server

# Or with Python
python -m mcp_postal_geocoder.server.mcp_server

CLI Client

# Test all tools
postal-cli test

# Individual commands
postal-cli geocode 90210
postal-cli reverse 47.606 -122.332
postal-cli stats

Streamlit Demo

streamlit run streamlit_app.py

🗂️ Database Schema

• zcta_code (TEXT) - 5-digit ZIP code (primary key)
• latitude, longitude (REAL) - WGS84 coordinates
• state (TEXT) - State abbreviation
• city (TEXT) - City name
• land_area_sqm, water_area_sqm (REAL) - Area measurements
• country_code (TEXT) - Always "US"

🔧 Development

# Install in development mode
pip install -e .

# Run tests
pytest

# Lint and format  
ruff check src/
black src/
mypy src/

📊 Performance

• Exact postal code lookup: <1ms
• Prefix searches: <10ms
• Reverse geocoding: <50ms
• Concurrent connections: 100+
• Database size: 6.9MB SQLite

🏗️ Architecture

src/mcp_postal_geocoder/
├── server/
│   ├── mcp_server.py          # Main MCP server (FastMCP)
│   ├── database/
│   │   ├── connection.py      # SQLite connection management
│   │   ├── models.py          # Pydantic data models
│   │   └── queries.py         # Optimized SQL queries
│   └── utils/
│       └── formatters.py      # Response formatting
├── client/
│   └── cli_tool.py           # CLI client for testing
└── data/
    └── postal_census_complete.db  # SQLite database (33,791 records)

🌐 API Compatibility

Designed as a drop-in replacement for GeoNames postal API:

{
  "totalResultsCount": 1,
  "geonames": [
    {
      "postalCode": "90210",
      "lat": 34.102512,
      "lng": -118.415075,
      "countryCode": "US", 
      "state": "CA",
      "placeName": "Santa Monica"
    }
  ]
}

📄 License

MIT License - see LICENSE file for details.

🚀 Deployment

HuggingFace Spaces

1. Create a new Space on HuggingFace with Streamlit SDK
2. Copy files: streamlit_app.py, requirements.txt, src/, data/
3. Use space_config.yml for Space configuration
4. Or use the GitHub Action for automatic sync

GitHub Action Setup

1. Add HF_TOKEN to GitHub repository secrets
2. Update USERNAME/SPACENAME in .github/workflows/sync-to-hf.yml
3. Push to main branch triggers automatic deployment

Local Development

streamlit run streamlit_app.py

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

📈 Status

• ✅ Phase 1: Core MCP server implementation
• ✅ Phase 2: Advanced search and city assignments
• ✅ Phase 3: CLI client and testing tools
• ✅ Phase 4: Streamlit demo and deployment
• 🔄 Phase 5: Documentation and optimization

🙏 Acknowledgments

• US Census Bureau for ZCTA data
• Anthropic for MCP framework
• FastMCP for modern server implementation