🚀 Deep Dive into Claude

A MCP (Model Context Protocol) server that extends the research capabilities of Claude and other MCP-compatible AI assistants. This server integrates web and academic search functions, enabling AI models to access current information from multiple sources, follow relevant links, and provide structured research results.

🚀 Quick Start

Overview

Deep Dive into Claude is a powerful research tool that extends the capabilities of LLMs in the following ways:

1. Web Search: Implemented through DuckDuckGo
2. Academic Research: Accessed via Semantic Scholar
3. Content Extraction: Extract relevant information from web pages
4. Comprehensive Analysis: Analyze using a structured format
5. Visualization Guidance: Provide guidance on data representation

This server adheres to MCP design principles and integrates seamlessly with Claude and other AI assistants.

✨ Features

• Unified Research Tool: Access web and academic information from a single interface
• Multi-source Integration: Integrate information from different sources into a coherent research
• Content Extraction: Extract relevant information from web pages
• Academic Resource Discovery: Find academic papers related to the topic
• Intelligent Formatting: Properly format research results
• Visualization Framework: Provide guidance on creating effective data visualizations
• Structured Analysis: Organize research using academic methods

📦 Installation

Prerequisites

• Python 3.8 or higher
• pip or uv package manager

Quick Installation

# Use pip
pip install mcp httpx beautifulsoup4

# Clone the repository
git clone https://github.com/yourusername/claude-deep-research.git

Configuration

The server can be used with default settings, but you can modify the following parameters in deep_research.py as needed:

# Configuration
USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36"
MAX_CONTENT_SIZE = 8000  # Maximum number of characters in the final response
MAX_RESULTS = 3         # Maximum number of results to process

💻 Usage Examples

Running the Server

Modify the Claude desktop configuration and restart Claude. On a Mac, the path is ~/Library/Application Support/Claude:

  "search-scholar": {
      "command": "<Python Path>/python",
      "args": [
        "<Deep Dive Path>/deep_research.py"
      ]
    }

Using with Claude Desktop

After installation, you can access the server in Claude Desktop:

1. Tool Access: Use the deep_research tool directly in the conversation

Research Tools

The main deep_research tool accepts the following parameters:

• query (required): Research question or topic
• sources (optional): Sources to use: "web", "academic", or both
• max_results (optional): Maximum number of results to return, default is 10

Example Prompts

# Query the latest research papers on artificial intelligence
deep_research query --query "latest AI research papers" --sources academic --max_results 5

# Search for relevant information on the web
deep_research search --query "AI applications in healthcare"

Troubleshooting

If you encounter problems, check the following:

• Ensure that Python and all dependencies are installed correctly
• Verify that the API key is set correctly
• Check the log files for more error information

Logging

Server output logs can be found in the logs/ directory. By default, verbose logging is enabled.

Guide to Deep Dive into Claude

Overview

Deep Dive into Claude is a powerful tool that helps you obtain and analyze information from web and academic resources. With simple commands, you can quickly get the results you need.

Installation and Configuration

Installation Steps

1. Install the necessary dependencies using pip:

pip install mcp httpx beautifulsoup4

2. Clone the repository to access the deep dive script:

git clone https://github.com/yourusername/claude-deep-research.git

3. Configure environment variables to ensure the server runs properly.

Environment Configuration

In deep_research.py, you may need to adjust the following settings:

# Set the request header to avoid anti-crawling mechanisms
USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36"

# Control the maximum content size returned
MAX_CONTENT_SIZE = 8000

# Set the maximum number of results to process
MAX_RESULTS = 3

Usage Guide

Starting the Server

Run the following command in the terminal to start the deep dive server:

python deep_research.py

Basic Commands

1. Query Academic Resources:

deep_research query --query "latest AI research papers" --sources academic --max_results 5

2. Search Web Information:

deep_research search --query "AI applications in healthcare"

Advanced Features

• Customize Request Headers: Modify the USER_AGENT or other request header parameters in deep_research.py.
• Limit Returned Results: Control the maximum number of returned results by setting MAX_RESULTS.

Troubleshooting

If you encounter problems, follow these steps to check:

1. Ensure that all dependencies are installed correctly.
2. Verify that the configuration file is correct.
3. Check the log files for detailed error information:

# View the log file
cat logs/deep_research.log

Logging

The server generates logs during operation, which are stored in the logs/ directory. By default, verbose logging is enabled, which helps with debugging and monitoring the server status.

This document provides you with a comprehensive guide to help you use the Deep Dive into Claude tool for efficient research. We hope this information is helpful to you!