Skip to main content
Glama
README.md4.36 kB
# MCP Reference Client This is a reference implementation of an MCP client that demonstrates how to connect to the MCP IDE Bridge messaging server for **client-to-client communication**. ## 🌟 Purpose This client serves as both a **testing tool** for the messaging server and a **reference implementation** for developers building their own MCP clients. It demonstrates: - **HTTP Streamable Transport**: Connects to MCP servers using the latest HTTP Streamable transport - **Client-to-Client Messaging**: Enables communication between different IDEs and tools - **Tool Execution**: Automatically executes MCP tools based on natural language queries - **Interactive Testing**: Provides a command-line interface for testing the messaging server ## 🎯 Use Cases ### IDE Collaboration Testing - Test messaging between different IDE instances - Verify cross-editor communication workflows - Debug MCP connection issues - Validate message routing and delivery ### Development Workflows - **Cursor ↔ Cursor**: Test pair programming scenarios - **Cursor ↔ VS Code**: Verify cross-editor file sharing - **Windsurf ↔ Any IDE**: Test AI agent coordination - **Team Coordination**: Simulate multi-developer workflows ## 🚀 Quick Start ### 1. Prerequisites - Python 3.12+ - MCP IDE Bridge server running on port 8111 - Optional: Anthropic API key for full functionality ### 2. Setup **Install dependencies:** ```bash pip install -r requirements.txt ``` **Set up environment (optional):** Create a `.env` file with your Anthropic API key for full functionality: ``` ANTHROPIC_API_KEY=your_api_key_here ``` **Note**: The client works without an API key for basic connection testing. ### 3. Usage **Start the messaging server first:** ```bash # In the main project directory python -m mcp_messaging.server --port 8111 ``` **Test basic connection (no API key required):** ```bash python test_connection.py --mcp-localhost-port 8111 ``` **Run full client (requires API key):** ```bash python client.py --mcp-localhost-port 8111 ``` ## 🧪 Testing Examples ### Connection Test ```bash python test_connection.py --mcp-localhost-port 8111 ``` This will: - Connect to the MCP server - List all available tools - Verify the connection is working ### Interactive Testing ```bash python client.py --mcp-localhost-port 8111 ``` **Example interactions:** ``` Query: Check in as client_1 with name "Test Client" Query: Send a message from client_1 to client_2 saying "Hello from client 1!" Query: Get messages for client_2 Query: What are the active sessions? ``` ## 🔧 How It Works 1. **Connection**: Establishes HTTP Streamable connection to the MCP server 2. **Tool Discovery**: Retrieves available tools from the server 3. **Query Processing**: Sends user queries to Claude with available tools (if API key provided) 4. **Tool Execution**: Executes MCP tools based on analysis 5. **Response Handling**: Processes tool results and continues conversation ## 🏗️ Architecture - **MCPClient**: Main client class handling MCP protocol communication - **Claude Integration**: Uses Anthropic's API for natural language processing (optional) - **Tool Management**: Automatically handles tool discovery and execution - **Session Management**: Proper cleanup of MCP sessions and streams ## 📋 Available Tools The client can interact with all MCP IDE Bridge tools: | Tool | Purpose | Example | |------|---------|---------| | `checkin_client` | Register presence | "Check in as alice_cursor" | | `send_message_without_waiting` | Send message (fire & forget) | "Send message to bob_vscode" | | `get_messages` | Retrieve messages | "Get messages for alice_cursor" | | `get_my_identity` | Get configuration help | "What's my identity?" | | `get_active_sessions` | View active connections | "Show active sessions" | ## 🔒 Security Note This client is designed for **testing and development** purposes. For production use, consider: - Using the enterprise solution at [MilesDyson.ai](https://milesdyson.ai) - Implementing proper authentication - Adding message encryption - Setting up audit logging ## 🤝 Contributing This reference client is part of the open source MCP IDE Bridge project. Contributions are welcome! - Report issues in the main repository - Submit pull requests for improvements - Help improve documentation and examples

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Mvp2o-ai/mcp-ide-bridge'

If you have feedback or need assistance with the MCP directory API, please join our Discord server