Spring AI MCP (Model Context Protocol) Server and Client

This repository demonstrates the use of the Model Context Protocol (MCP) with Spring AI. MCP is a protocol for structuring, exchanging, and managing context between clients, servers, and AI models, enabling advanced AI-powered applications.

It contains two Spring Boot applications:

• spring-ai-mcp-server: A backend RESTful API server implementing MCP for managing medical reports and patient data, and orchestrating communication with AI models.
• spring-ai-mcp-client: A client application that interacts with the server using MCP, providing a user interface and consuming the server's AI-powered APIs.

Both modules are built with Java, Spring Boot, and Gradle, and leverage Spring AI's MCP integration for seamless context management and AI model interaction.

---

Table of Contents

• Architecture Overview
• Server Module
  • Features
  • Project Structure
  • Key Classes
  • Configuration
• Client Module
  • Features
  • Project Structure
  • Key Classes
  • Configuration
• Build & Run
• API Endpoints
• License

---

Architecture Overview

This system is designed around the Model Context Protocol (MCP), which standardizes how context is exchanged between the client, server, and AI model. MCP enables:

• Consistent context passing for AI inference
• Enhanced traceability and explainability
• Modular integration of different AI models (e.g., Llama via Ollama)

Server: Implements MCP endpoints, manages business logic, and communicates with the AI model using Spring AI's MCP support.

Client: Consumes MCP endpoints, structures requests using the protocol, and presents results to users or other systems.

---

Server Module

Features

• Implements Model Context Protocol (MCP) for context-rich AI interactions
• RESTful API for managing medical reports and patients
• Layered architecture: Controller, Service, Repository, Entity, DTO
• Data persistence (typically with JPA/Hibernate)
• Exception handling
• Configurable via application.yaml

Project Structure

src/main/java/com/spring/ai/mcp/server/
  ├── SpringAiMcpServerApplication.java
  ├── controller/           # (if present) REST controllers
  ├── dto/                  # Data Transfer Objects
  ├── entity/               # JPA Entities
  ├── repository/           # Spring Data JPA Repositories
  └── service/              # Business logic
src/main/resources/
  └── application.yaml      # Configuration

Key Classes

• SpringAiMcpServerApplication.java: Main entry point for the server application.
• entity/: Contains JPA entities like Patient and MedicalReport.
• dto/: Contains DTOs for API requests/responses, e.g., PatientDTO, MedicalReportDTO.
• repository/: Spring Data interfaces for database access.
• service/MedicalService.java: Business logic for managing patients and reports.
• MCP Integration: Uses Spring AI's MCP modules to structure and manage context for AI model calls.

Configuration

• src/main/resources/application.yaml: Contains server configuration (port, DB, etc).

---

Client Module

Features

• Consumes MCP-enabled REST APIs from the server
• Structures requests and responses using Model Context Protocol
• May provide a web UI or further REST endpoints
• Configurable via application.yaml

Project Structure

src/main/java/com/spring/ai/mcp/client/
  ├── SpringAiMcpClientApplication.java
  └── controller/           # Controllers for client endpoints/UI
src/main/resources/
  └── application.yaml      # Configuration

Key Classes

• SpringAiMcpClientApplication.java: Main entry point for the client application.
• controller/MedicalController.java: Handles client-side endpoints or UI logic.
• MCP Integration: Uses Spring AI's MCP client modules to structure and send context-rich requests to the server.

Configuration

• src/main/resources/application.yaml: Contains client configuration (server URL, etc).

---

Build & Run

Prerequisites

• Java 17+
• Gradle 7+

Build

From each module directory (spring-ai-mcp-server and spring-ai-mcp-client):

./gradlew build

Run

From each module directory:

./gradlew bootRun

---

API Endpoints

MCP Server Endpoints (examples)

• GET /api/patients - List all patients
• POST /api/patients - Create a new patient
• GET /api/reports - List all medical reports
• POST /api/reports - Create a new medical report
• POST /api/ai/infer - Send context to AI model using MCP (example)

(See controller classes and Spring AI MCP documentation for full details.)

---

License

This project is licensed under the MIT License.