Where's my train? MCP Server

A MCP server for real-time NYC subway information. Leverages AI/LLM geographic knowledge for intelligent location handling and provides structured JSON data for stations, arrivals, alerts, and service status with live MTA data.

Data from MTA • Personal use only • Not endorsed by MTA • Data provided "as is"

Quick Examples

• "where's my train? I'm looking for the Q at DeKalb Av"

Features

• Real-time train arrivals with crowding information

• Station search with fuzzy matching and accessibility info

• Service alerts and disruption analysis

• Intelligent location handling - AI converts location names to coordinates

• Nearby station finder using GPS coordinates or location names

• Transfer information for complex routes

• Live MTA GTFS-RT data from all subway lines

Usage

Option 1: Use on Smithery (Recommended)

Instant access - Visit smithery.ai/server/@sasabasara/where_is_my_train_mcp and click "Install" to add it to your AI client.

No setup required - works immediately with Claude, ChatGPT, Cursor, and other MCP-compatible clients.

Option 2: Run Locally

Prerequisites:

• Node.js 18+

Quick Start:

That's it! The server automatically downloads and caches MTA GTFS data on first run.

What You Can Do

Station-Based Queries:

• Find stations: "find Times Square station"

• Next trains: "next trains at Union Square"

• Transfers: "transfers at Atlantic Ave"

System Information:

• Service status and alerts

• Real-time disruptions

• Train crowding data

Location-Based Searches:

• Nearest stations: "stations near Times Square" or "near SoHo"

• GPS coordinates: lat: 40.7589, lon: -73.9851

• AI automatically converts location names to coordinates

What You Cannot Do (MTA Data Limitations)

Complex Journey Planning:

• ❌ Multi-modal transit (bus + subway combinations)

• ❌ Real-time traffic-aware routing

• ❌ Ride-sharing integration

• ✅ AI can chain tools for basic subway journey planning

Non-Subway Transit:

• ❌ Bus routes and schedules

• ❌ LIRR, Metro-North, NJ Transit

• ❌ Ferry, taxi, rideshare information

• ✅ NYC Subway only (all lines: 1-7, A-Z, shuttles)

Historical/Future Data:

• ❌ Past service performance

• ❌ Schedules beyond ~2 hours

• ❌ Planned service changes (beyond current alerts)

• ✅ Real-time data only (current conditions)

Tools

Core Tools

• next_trains - Real-time arrivals with crowding indicators

• find_station - Fuzzy station search with accessibility info

• nearest_station - Find closest stations (AI converts location names to coordinates)

• station_transfers - Transfer options at stations

System Tools

• service_status - System-wide or line-specific status

• subway_alerts - Detailed service alerts with filtering

• service_disruptions - Disruption analysis with alternatives

Example Queries

MTA Compliance & Usage Terms

Personal Use Only - This server is designed for individual, non-commercial use.

Key Requirements:

• ✅ Personal development - Direct MTA feed access permitted

• ❌ Public distribution - Requires data caching server and MTA license

• 📋 Attribution required - "Data from MTA" in outputs

• 🚫 No redistribution of raw feed data

Data Disclaimers:

• Data provided "as is" without accuracy guarantees

• May be delayed or incomplete due to processing

• Not affiliated with or endorsed by MTA

For production/public use, you must implement proper data caching infrastructure and obtain MTA licensing.

Deployment

Smithery (Recommended)

Traditional MCP

Use npm start for stdio protocol with any MCP client.

Data Sources

• Real-time feeds: All NYC subway lines via MTA GTFS-RT

• Static data: Stations, routes, transfers (auto-downloaded)

• Service alerts: Live disruption and delay information

• Update frequency: Every 30 seconds (real-time data)

License

MIT License - see LICENSE file for details