supermanzer/mcp_client_sever

MCP Research Assistant

An intelligent research assistant that combines the Model Context Protocol (MCP) with Anthropic's Claude API to help researchers search, organize, and analyze academic papers from arXiv.

Features

• Interactive Chat Interface: Natural language interaction with the research assistant
• arXiv Integration: Search and retrieve papers directly from arXiv
• Topic Organization: Papers are automatically organized by research topics
• Metadata Storage: Structured storage of paper information in JSON format
• Research Summaries: Generate comprehensive research summaries across multiple papers
• Multi-Server Architecture: Supports multiple MCP servers with different tool sets
• Asynchronous Operation: Efficient handling of concurrent operations

Requirements

• Python 3.7+
• Anthropic API key (stored in .env file)
• server_config.json for MCP server configurations
• Required Python packages (see Dependencies section)

Installation

1. Clone the repository
2. Create a .env file with your Anthropic API key:

   API_KEY=your_anthropic_api_key_here
   

3. Create a server_config.json file with your MCP server configurations
4. Install the required dependencies

Dependencies

• anthropic: Anthropic Claude API client
• arxiv: arXiv API client
• mcp: Model Context Protocol implementation
• python-dotenv: Environment variable management
• nest-asyncio: Asyncio support in interactive environments

Project Structure

├── client.py           # MCP chatbot client implementation
├── server.py           # arXiv research paper management server
├── server_config.json  # Server configuration file
└── papers/            # Directory for storing paper information
    └── topic_name/    # Topic-specific directories
        └── papers_info.json  # Paper metadata storage

Usage

1. Start the client (the server will start automatically):

   python client.py
   

2. Interact with the chatbot using natural language queries. Examples:

   • "Search for papers about quantum computing"
   • "Show me available research topics"
   • "Get information about paper [paper_id]"
   • "Summarize recent papers in machine learning"

Available Tools

Search Papers

• Search for papers on arXiv by topic
• Automatically store paper metadata
• Organize papers by research topics

Extract Information

• Retrieve detailed information about specific papers
• Access paper metadata across all topics
• Get formatted paper summaries

Browse Topics

• List available research topics
• View papers within each topic
• Access topic-specific summaries

Features in Detail

Paper Information Storage

Each paper's metadata includes:

• Title
• Authors
• Summary
• PDF URL
• Publication date

Research Summaries

Generated summaries include:

• Overview of current research state
• Common themes and trends
• Key research gaps
• Impactful papers
• Methodologies used

Error Handling

• Graceful handling of API errors
• Corruption detection in storage files
• Connection error management
• Automatic resource cleanup

Architecture

The project uses a client-server architecture:

Server (server.py)

• Implements MCP server functionality
• Manages paper storage and organization
• Provides tools for paper search and analysis
• Handles arXiv integration

Client (client.py)

• Implements interactive chat interface
• Manages connections to MCP servers
• Integrates with Claude AI
• Coordinates tool usage
• Handles user interaction

Contributing

Feel free to submit issues, fork the repository, and create pull requests for any improvements.

License

[Specify your license here]