MCP Chat Application

This is an interactive chat application that uses the Model Control Protocol (MCP) with Groq's language models. The application provides a command-line interface for having conversations with an AI assistant while maintaining conversation memory.

Features

• Interactive command-line chat interface
• Integration with Groq's language models (using qwen-qwq-32b)
• Conversation memory management
• Simple commands for clearing history and exiting
• Browser automation capabilities through MCP

Prerequisites

• Python 3.x
• Required API keys:
  • GROQ_API_KEY

Installation

1. Clone this repository:

git clone <https://github.com/rohith372/MCP>
cd <https://github.com/rohith372/MCP>

2. Install the required dependencies:

pip install -r requirements.txt
uv add langchain-groq
uv add langchain-openai
uv add mcp-use

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

GROQ_API_KEY=your_groq_api_key_here

Usage

1. Run the chat application:

uv run app.py

2. Available commands during chat:

• Type your message and press Enter to chat with the assistant
• Type clear to clear the conversation history
• Type exit or quit to end the conversation

Configuration

The application uses a configuration file browser_mcp.json for MCP client settings. Make sure this file is present in your project directory with the appropriate configuration.

Dependencies

• langchain-groq
• python-dotenv
• mcp-use
• asyncio

Error Handling

The application includes basic error handling for:

• API communication issues
• Invalid inputs
• Session management