A Spring Boot web service that provides a REST API for interacting with AI language models through Ollama and Model Context Protocol (MCP) tools.

This application serves as a bridge between web clients and AI capabilities, combining:

• Ollama AI Chat Model: Local AI model integration using Llama 3.1 8B
• MCP (Model Context Protocol) Integration: Tool execution through MCP servers
• REST API: HTTP endpoints for chat interactions and system monitoring

• Simple Chat: Direct conversation with the AI model
• Conversational Chat: Maintains conversation history with unique session IDs
• System Prompt Chat: Custom system instructions for specialized responses
• Smart Tool Detection: Automatically determines when to use MCP tools based on user input

• Dynamic Tool Loading: Automatically discovers and loads MCP server tools
• GitHub Tools: Integrated MCP server for GitHub operations
• Builder Tools: Integrated MCP server for build operations
• Tool Status Monitoring: Real-time status of connected MCP servers

• Health Checks: Monitor Ollama and MCP server connectivity
• Tool Discovery: List available tools from all connected MCP servers
• Server Status: Track connection status of individual MCP servers
• Actuator Endpoints: Spring Boot management endpoints for monitoring

• POST /api/chat - Simple chat interaction
• POST /api/chat/conversation - Chat with conversation history
• POST /api/chat/system - Chat with custom system prompt
• DELETE /api/chat/conversation/{id} - Clear conversation history

• GET /api/health - Application and service health status
• GET /api/info - Service information and configuration
• GET /api/tools - List available MCP tools
• GET /api/mcp/status - MCP server connection status

The application is configured via application.yaml:

• Model: Llama 3.1 8B
• Base URL: http://localhost:11434
• Timeout: 300 seconds

• GitHub Server: ./servers/mcp-server-github-0.0.1-SNAPSHOT.jar
• Builder Server: ./servers/mcp-server-builder-0.0.1-SNAPSHOT.jar
• Request Timeout: 60 seconds

• Java 21
• Maven 3.6+
• Ollama running locally on port 11434 with Llama 3.1 8B model
• MCP Server JARs placed in ./servers/ directory

1. Install Ollama and pull the required model:

   ollama pull llama3.1:8b

2. Create servers directory and add MCP server JARs:

   mkdir servers
   # Copy your MCP server JARs to ./servers/

3. Build and run:

   mvn clean package
   java -jar target/mcp-client-assistant-0.0.1-SNAPSHOT.jar

4. Access the service at http://localhost:8080

curl -X POST http://localhost:8080/api/chat \
  -H "Content-Type: application/json" \
  -d '{"message": "Hello, how are you?"}'

curl -X POST http://localhost:8080/api/chat \
  -H "Content-Type: application/json" \
  -d '{"message": "List files in the current directory"}'

curl http://localhost:8080/api/health

• Spring Boot 3.5.5 with Java 21
• Spring AI for LLM integration and MCP client support
• Ollama Integration for local AI model execution
• MCP Client for tool execution via external servers
• RESTful API design with JSON request/response
• Conversation Management with in-memory session storage

This project is licensed under the terms specified in the LICENSE file.