Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to content

Latest commit

 

History

History

monkey-mcp-client

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
.mvn/wrapper
.mvn/wrapper
 
 
src/main
src/main
 
 
README.md
README.md
 
 
mcp.json
mcp.json
 
 
mvnw
mvnw
 
 
mvnw.cmd
mvnw.cmd
 
 
pom.xml
pom.xml
 
 

README.md

Monkey MCP Console Application

A Quarkus console application that integrates with AI models through Ollama and registers MCP (Model Context Protocol) servers for additional tools and capabilities.

Features

  • Chat Interface: Interactive console chat with Ollama LLMs
  • MCP Integration: Connect to the Monkey species MCP server via HTTP SSE transport
  • Tool Discovery: List and use available tools from MCP servers
  • Modular Design: Clean separation of concerns with services and commands

Technology Stack

  • Java 21 with Maven
  • Picocli for console application interface
  • LangChain4J for AI chat interactions and MCP client support
  • Ollama for local LLM hosting
  • Jackson for JSON processing

Prerequisites

  1. Java 21 or later
  2. Ollama running locally on http://localhost:11434 with llama3.2 model
  3. Monkey MCP Server running on http://localhost:8080/mcp/sse

Quick Start

1. Install Ollama and pull the model

# Install Ollama (macOS)
brew install ollama

# Start Ollama service
ollama serve

# Pull the llama3.2 model
ollama pull llama3.2

2. Start the Monkey MCP Server

Ensure the monkey species MCP server is running at http://localhost:8080/mcp/sse.

3. Build the project

# Build
./mvnw verify

Available Commands

Help Command

Show available commands:

java -jar target/cli-monkey-mcp-client.jar -h

Tools Command

List all available MCP tools from registered servers:

java -jar target/cli-monkey-mcp-client.jar tools

Chat Command

Start an interactive chat session with the Ollama LLM:

java -jar target/cli-monkey-mcp-client.jar chat

MCP Configuration

The application connects to MCP servers defined in mcp.json file located where the application is launched from (working directory):

{
  "servers": {
    "monkey-species": {
      "type": "sse",
      "url": "http://localhost:8080/mcp/sse"
    }
  }
}

Project Structure

src/main/java/org/acme/
├── McpClientApplication.java    # Main application class
├── client/
│   ├── ChatService.java         # Ollama chat service
│   └── ToolsService.java        # MCP client service
└── command/
    ├── ChatCommand.java         # Chat command implementation
    └── ToolsCommand.java        # Tools command implementation

Example Usage

Chat Session

$ java -jar target/cli-monkey-mcp-client.jar chat

═══════════════════════════════════
      Java MCP Chat Client
═══════════════════════════════════
→ Starting chat with local Ollama (llama3.2) + MCP Tools
→ Available tools: 2 monkey species tools

💡 Try asking:
   • 'What monkey species do you know?'
   • 'Tell me about a random monkey'

Type 'exit', 'quit', or 'bye' to end the conversation
───────────────────────────────────

You: Tell me about capuchin monkeys
AI: I can help you with monkey species information! I have access to MCP tools that can provide details about various monkey species, their habitats, and characteristics...

You: List all monkey species
AI: I have access to several MCP tools for monkey species information:
- get_monkey_species: Get detailed information about a specific species
- list_monkey_species: List all available monkey species
- get_monkey_habitat: Get habitat information for a species

You: exit
👋 Goodbye!

Tools Listing

$ java -jar target/cli-monkey-mcp-client.jar tools
...

═════ MCP Tools ═════

Tool: get_monkey_species_details
  Description: Get detailed information for a specific monkey species by name
  Parameters: JsonObjectSchema {description = null, properties = {speciesName=JsonStringSchema {description = null }}, required = [speciesName], additionalProperties = null, definitions = {} }

Tool: get_monkey_species_stats
  Description: Get statistics about the monkey species database
  Parameters: JsonObjectSchema {description = null, properties = {}, required = [], additionalProperties = null, definitions = {} }

Tool: get_random_monkey_species
  Description: Get a random monkey species with full details
  Parameters: JsonObjectSchema {description = null, properties = {}, required = [], additionalProperties = null, definitions = {} }

Tool: list_monkey_species
  Description: List all available monkey species with their basic information
  Parameters: JsonObjectSchema {description = null, properties = {}, required = [], additionalProperties = null, definitions = {} }

───────────────────────────────────
Total: 4 tools available
→ Shutting down MCP tools service...
✓ MCP client monkeymcp closed
✓ All MCP clients closed

Troubleshooting

  1. Ollama Connection Issues: Ensure Ollama is running on localhost:11434
  2. MCP Server Connection: Verify the monkey MCP server is accessible at localhost:8080/mcp/sse
  3. Model Not Found: Pull the required model with ollama pull llama3.2