Spring MCP Server

The Model Context Protocol is an open standard that enables developers to build secure, two-way connections between their data sources and AI-powered tools. The architecture is straightforward: developers can either expose their data through MCP servers or build AI applications (MCP clients) that connect to these servers.

Prerequisites

• Java 21
• Maven 3.9.9
• Spring Boot 3.4.4
• Spring AI 0.4.0
• Spring Model Context Protocol Server
• Claude Desktop Application

Getting Started

In this example, we will create a simple Spring Boot application that serves as an MCP server.

1. Create a Spring Boot Application

Create a new Spring Boot application using the Spring Initializr or your preferred method. You can use the following dependencies:

• Spring Model Context Protocol Server

Creating record

Create a record class that will represent the information you want to expose through the MCP server.

package org.sehn.spring_mcp;

public record Info(String title, String url) {
}

Creating services

Create a service class that will provide the information you want to expose through the MCP server.

package org.sehn.spring_mcp;

import jakarta.annotation.PostConstruct;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.ai.tool.annotation.Tool;
import org.springframework.stereotype.Service;

import java.util.ArrayList;
import java.util.List;

@Service
public class InfoService {

    private static final Logger logger = LoggerFactory.getLogger(InfoService.class);

    private List<Info> information = new ArrayList<>();

    @Tool(name = "dr_get_information", description = "Get a list of information about Deepanshu Rawat")
    public List<Info> getInformation() {
        return information;
    }

    @Tool(name = "dr_get_info", description = "Get a specific information about Deepanshu Rawat")
    public Info getInfo(String title) {
        return information.stream().filter(info -> info.title().equals(title)).findFirst().orElse(null);
    }

    @PostConstruct
    public void init() {
        information.addAll(List.of(
                new Info("Portfolio", "https://bento.me/deepanshu-rawat6"),
                new Info("GitHub", "https://github.com/deepanshu-rawat6"),
                new Info("LinkedIn", "https://www.linkedin.com/in/deepanshu-rawat6/"),
                new Info("Twitter", "https://twitter.com/deepanshuurawat")
        ));


    }
}

Creating a Tool Callback

In the SpringMCPApplication or your main application class, create a tool callback that will handle the requests from the MCP server.

    @Bean
    public List<ToolCallback> deepanshuTools(InfoService infoService) {
        return List.of(ToolCallbacks.from(infoService));
    }

Adding Configuration

Create a configuration class that will configure the MCP server.

spring.application.name=spring_mcp
spring.main.web-application-type=none
spring.ai.mcp.server.name=spring_mcp
spring.ai.mcp.server.version=0.0.1


spring.main.banner-mode=off
logging.pattern.console=

2. Build the Application

Simply run the following command to build the application:

mvn clean package -DskipTests

3. Further configuration

Now, in order to use the MCP server, you need to add the following configuration in your claude_desktop_config.json:

{
    "mcpServers": {
        "<name_of_the_mcp_server>": {
            "command": "java",
            "args": [
                "-jar",
                "<Path_of_the_jar_file_created_after_building>"
            ]
        }
    }
}

Replace <name_of_the_mcp_server> with the name of your MCP server and <Path_of_the_jar_file_created_after_building> with the path to the jar file created after building the application.

For example, name of my MCP server is spring_mcp.

4. Testing the MCP Server

Now, start Claude Desktop and a few new things will appear underneath the prompt area:

Installed MCP Servers

Available MCP tools

Now, upon hitting a prompt, say for example: Get me the information about Deepanshu Rawat, you will get a alert about third-party tools being used, simply accept the alert for the current chat. And then you will get a response, like this: