MCP Client

MCP Client

A TypeScript client for the Model Context Protocol (MCP) that enables communication between language models and external MCP server tools/services.

Overview

This MCP client provides a flexible interface for connecting to various MCP servers, allowing language models like OpenAI's GPT and Anthropic's Claude to access external tools and services through a standardized protocol. The client supports multiple server types and handles cross-platform compatibility.

Features

• Connect to multiple MCP servers simultaneously
• Support for both OpenAI and Anthropic language models
• Cross-platform compatibility (Windows, macOS, Linux)
• Automatic handling of platform-specific commands
• Configurable environment variables

Prerequisites

• Node.js (v16+)
• npm or yarn
• TypeScript

Installation

1. Clone the repository:

   git clone https://github.com/kamusis/mcp-client.git
   cd mcp-client
   

2. Install dependencies:

   npm install
   

3. Create a .env file in the project root with your API keys:

   # LLM (OPENAI or ANTHROPIC)
   LLM=OPENAI
   OPENAI_MODEL=gpt-4o
   ANTHROPIC_MODEL=claude-3-5-sonnet-20241022
   
   # LLM API Key
   ANTHROPIC_API_KEY=your_anthropic_api_key
   OPENAI_API_KEY=your_openai_api_key
   
   # MCP Configuration File Path (optional, defaults to ./mcp_config.json)
   # MCP_CONFIG_PATH=./mcp_config.json
   

4. Configure your MCP servers in mcp_config.json:

   {
     "mcpServers": {
       "sequential-thinking": {
         "command": "npx",
         "args": [
           "-y",
           "@modelcontextprotocol/server-sequential-thinking"
         ]
       },
       "postgres-full-localhost-dev": {
         "command": "node",
         "args": [
           "/path/to/mcp-postgres-full-access-extended/dist/index.js",
           "postgresql://postgres:[email protected]:5432/postgres"
         ],
         "env": {
           "TRANSACTION_TIMEOUT_MS": "60000",
           "MAX_CONCURRENT_TRANSACTIONS": "20",
           "PG_STATEMENT_TIMEOUT_MS": "30000"
         }
       }
     }
   }
   

Building and Running

1. Build the project:

   npm run build
   

2. Run the client:

   node build/index.js
   

Configuration

Environment Variables

• LLM: The language model provider to use (OPENAI or ANTHROPIC)
• OPENAI_MODEL: The OpenAI model to use (e.g., gpt-4o)
• ANTHROPIC_MODEL: The Anthropic model to use (e.g., claude-3-5-sonnet-20241022)
• ANTHROPIC_API_KEY: Your Anthropic API key
• OPENAI_API_KEY: Your OpenAI API key
• MCP_CONFIG_PATH: Path to the MCP configuration file (optional, defaults to ./mcp_config.json)

MCP Server Configuration

The mcp_config.json file defines the MCP servers to connect to. Each server entry includes:

• command: The command to run the server
• args: Arguments to pass to the command
• env: Environment variables for the server (optional)

Cross-Platform Compatibility

The client automatically handles platform-specific differences:

• On Windows, npx commands are adjusted to use npx.cmd
• Python commands are adjusted based on the platform (python on Windows, python3 on Unix-like systems)
• Environment variables are properly passed to child processes

License

MIT