Spring MCP Server Demo

A simple demonstration of how to build a remote Model Context Protocol (MCP) server using Spring Boot and Java. This project showcases how to implement and expose tools that can be called by AI models through the MCP protocol.

What's Inside

• Complete MCP server implementation using Spring AI
• Tool registration and callback handling
• RAG search service as a practical example tool
• Docker support for easy deployment

Quick Start

# Run with Maven
mvn spring-boot:run

# Or build and run the JAR
mvn clean package
java -jar target/mcpserver-0.0.1-SNAPSHOT.jar

The server will start at http://localhost:8080

🧪 Testing with MCP Inspector

The MCP Inspector is a tool that allows you to test your MCP server without needing to integrate with an LLM.

# Run the Inspector with your MCP server
npx @modelcontextprotocol/inspector

This will start the MCP Inspector web interface (typically at http://127.0.0.1:6274)

Docker Support

Want to run this in Docker? No problem!

# Build and package everything in one go
./docker-build.sh

# Run the container
docker run -p 8080:8080 mcp-server:latest

That's it! Your MCP server will be available at http://localhost:8080.

Available Tools

Search

Searches through documentation files in the RAG directory.

Parameters:

• query: The search term

Returns: A list of relevant text snippets from the documentation

How It Works

This demo uses Spring AI's MCP server capabilities to expose tools that can be called by AI models. The server implements the Model Context Protocol, allowing AI models to discover and use the tools provided by this server.

The RAG search tool demonstrates how to implement a practical tool that can enhance an AI model's capabilities by providing access to specific information sources.