MCP 服务器

mcp-stm-montevideo

MCP server exposing Montevideo public transportation data (STM) as tools for AI assistants, enabling natural language queries about routes, stops, arrivals, and trip planning.

README

MCP STM Montevideo

MCP server exposing Montevideo public transportation data (STM) as tools for AI assistants.

This project allows AI agents and LLM-based applications to query public transport information such as bus routes, stops, arrivals, and connections in Montevideo through the Model Context Protocol (MCP).

The goal is to make city infrastructure data accessible through conversational interfaces.

Demo

https://github.com/user-attachments/assets/805a692b-b2cc-4223-9abf-e7d5edf99eb6

Features

Exposes Montevideo STM transport data as MCP tools
Supports natural language queries about routes, stops, arrivals, and trip planning
Designed for AI assistants such as Claude Desktop, Cursor, and other MCP clients
Includes a REST API layer in addition to MCP
Built with Node.js and TypeScript
Integrates public STM datasets into a developer-friendly interface

Example

User query

How do I go from Facultad de Ingenieria to Plaza Independencia?

Assistant response

Take a bus from the stops near Bv. Espana and continue toward Ciudad Vieja.
Get off near Plaza Independencia.

Architecture

The server exposes STM transport data through MCP tools that AI assistants can call while answering user requests.

AI Assistant
     |
     v
MCP Client
     |
     v
MCP STM Montevideo Server
     |
     v
STM Transport Data

Installation

Clone the repository:

git clone https://github.com/chaba11/mcp-stm-montevideo
cd mcp-stm-montevideo

Install dependencies:

npm install

Build the project:

npm run build

Run the MCP server:

npm run start

Run the REST API locally:

npm run dev:api

Example MCP Tools

Example tools exposed by the server:

buscar_parada
proximos_buses
recorrido_linea
ubicacion_bus
como_llegar

These tools allow AI assistants to retrieve structured transportation data and generate natural language responses for users.

Use Cases

AI assistants answering public transport questions
Conversational city navigation tools
Smart travel assistants
Urban mobility integrations for LLM applications
MCP and API-based transit experiences

Tech Stack

Node.js
TypeScript
MCP (Model Context Protocol)
Hono
OpenAPI / Swagger
Public STM transport data

Why this project

As AI assistants become more common, exposing real-world systems through MCP servers enables natural language interaction with infrastructure and public services.

This project explores how public transportation systems can integrate with the AI tooling ecosystem in a practical, developer-friendly way.

This project was also an experiment: exploring MCPs as a way to connect real-world data with LLMs, and evaluating autonomous software development — most of the code was generated with Claude Code following a methodology of sequential loops (Ralph Loops).