Example MCP Server + Client Implementation

This project demonstrates the Model-Context-Protocol (MCP) pattern, as explained in this article and this video.

It includes a backend service, an MCP server exposing the service, and an MCP client (an AI chatbot). The chatbot uses the MCP server to store and retrieve custom information via the backend service.

Important: This project requires Node.js 23.x or higher to execute TypeScript directly.

Configuration

Create a .env file in the client folder and add your OpenAI API key:

OPENAI_API_KEY=<your-open-ai-key>

Run npm install in the service, server, and client folders to install dependencies.

Running Service, MCP Server & MCP Client

In each folder (service, server, client), run npm run dev using Node.js 23+. The client is an AI chatbot that leverages the custom MCP server to interact with the backend service. This setup allows the AI chatbot to store and retrieve information.