Data Generator MCP Server Java

🚀 Data Generator Java MCP Server

This is a Spring Boot-based application that implements a Model Completion Protocol (MCP) server for data generation. This project demonstrates how to use Spring AI to create a custom MCP server and integrate it with an AI model to generate data based on prompts.

📋 Overview

This application runs as an MCP server, providing data generation tools. It leverages the MCP server capabilities of Spring AI to expose Java methods as tools for AI models to call. The server is configured to run in synchronous mode.

✨ Features

• Implement a Model Completion Protocol (MCP) server
• Provide data generation tools
• Support synchronous execution mode
• Include a REST client for external API integration
• Implement a simple user data model

📋 Prerequisites

• Java 21 or higher
• Maven 3.9+ (or use the included Maven wrapper script)
• Basic understanding of Spring Boot and Spring AI

🚀 Quick Start

🔨 Build the Application

1. Clone the repository:

   git clone <repository-url>
   cd data-generator-java-mcp-server
   

2. Build the application using Maven:

   ./mvnw clean install
   

▶️ Run the Application

Run the application using the Maven wrapper script:

./mvnw spring-boot:run

The MCP server will start and be ready to accept requests.

⚙️ Configuration

🛠️ MCP Server Configuration

The MCP server is configured in the application.properties file:

# Application name
spring.application.name=data-generator-java-mcp-server

# MCP server name
spring.ai.mcp.server.name=data-generator-java-mcp-server

# MCP server type (SYNC or ASYNC)
spring.ai.mcp.server.type=SYNC

# Disable Spring Boot banner
spring.main.banner-mode=off

# Set as non-web application
spring.main.web-application-type=none

⚙️ Custom Configuration

Customize the MCP server by modifying the McpServerConfig class:

@Configuration
public class McpServerConfig {

    @Bean
    public ToolCallbackProvider dataGeneratorToolCallbackProvider(DataGeneratorService dataGeneratorService) {
        return MethodToolCallbackProvider.builder()
                .toolObjects(dataGeneratorService)
                .build();
    }
}

💻 Available Tools

The MCP server exposes the following tools:

1. generateData - Generate data based on the provided prompt

   • Parameter: prompt (String)
   • Return: Generated data as a String

2. getUserData - Generate user data based on the provided ID and age

   • Parameters:
     • id (String)
     • age (int)
   • Return: A User object containing ID, name, and age

🚀 Extend the Application

To add new tools to the MCP server:

1. Create a new service class or extend the existing DataGeneratorService.
2. Add methods annotated with @Tool.
3. Register the service in the McpServerConfig class.

Example:

@Service
public class MyCustomService {

    @Tool(
        name = "myCustomTool",
        description = "Description of the tool"
    )
    public SomeReturnType myCustomTool(String param1, int param2) {
        // Implementation code
        return result;
    }
}

Then register it in the configuration:

@Bean
public ToolCallbackProvider myCustomToolCallbackProvider(MyCustomService myCustomService) {
    return MethodToolCallbackProvider.builder()
            .toolObjects(myCustomService)
            .build();
}

🔌 Integration with AI Models

To use the MCP server in conjunction with an AI model, the AI model can call the data generation methods by sending HTTP requests to the MCP server's API endpoints. The requests need to include the necessary parameters and authorization information.

In this way, the MCP server can serve as a core service for data generation, used by multiple AI applications or models.