paper-search-mcp 0.1.0

A MCP server for searching and downloading academic papers from multiple sources.

Paper Search MCP

A Model Context Protocol (MCP) server for searching and downloading academic papers from multiple sources, including arXiv, PubMed, bioRxiv, and Sci-Hub (optional). Designed for seamless integration with large language models like Claude Desktop.

---

Table of Contents

• Overview
• Features
• Installation
  • Quick Start
  • For Development
• Testing
• Usage
  • Running the Server
  • Using with Claude Desktop
  • Example Commands
• Deployment
  • Local Deployment
  • Cloud Deployment
• Contributing
• License

---

Overview

paper-search-mcp is a Python-based MCP server that enables users to search and download academic papers from various platforms. It provides tools for searching papers (e.g., search_arxiv) and downloading PDFs (e.g., download_arxiv), making it ideal for researchers and AI-driven workflows. Built with the MCP Python SDK, it integrates seamlessly with LLM clients like Claude Desktop.

---

Features

• Multi-Source Support: Search and download papers from arXiv, PubMed, bioRxiv, and Sci-Hub (optional).
• Standardized Output: Papers are returned in a consistent dictionary format via the Paper class.
• Asynchronous Tools: Efficiently handles network requests using httpx.
• MCP Integration: Compatible with MCP clients for LLM context enhancement.
• Extensible Design: Easily add new academic platforms by extending the academic_platforms module.

---

Installation

paper-search-mcp can be installed using uv or pip. Below are two approaches: a quick start for immediate use and a detailed setup for development.

Quick Start

For users who want to quickly run the server:

1. Install with uv (recommended):

   uv add paper-search-mcp
   

   OR with pip:

   pip install paper-search-mcp
   

2. Run the Server:

   uv run -m paper_search_mcp.server  # With uv
   # OR
   python -m paper_search_mcp.server  # With pip
   

3. Verify Installation:

   python -c "import paper_search_mcp; print(paper_search_mcp.__version__)"
   

For Development

For developers who want to modify the code or contribute:

1. Setup Environment:

   # Install uv if not installed
   curl -LsSf https://astral.sh/uv/install.sh | sh
   
   # Clone repository
   git clone https://github.com/yourusername/paper-search-mcp.git
   cd paper-search-mcp
   
   # Create and activate virtual environment
   uv venv
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate
   

2. Install Dependencies:

   # Install project in editable mode
   uv add -e .
   
   # Add development dependencies (optional)
   uv add pytest flake8
   

3. Verify Setup:

   # Check installation
   python -c "import paper_search_mcp; print(paper_search_mcp.__version__)"
   
   # Run tests
   python -m unittest discover tests
   

---

Testing

To ensure the server works correctly, run the included unit tests.

1. Activate Environment:

   source .venv/bin/activate
   

2. Run Tests:

   python -m unittest discover tests
   

   • test_arxiv.py: Tests the ArxivSearcher class.
   • test_server.py: Tests the MCP server tools (search_arxiv and download_arxiv).

   Expected Output:

   ....
   Ran 4 tests in X.XXXs
   OK
   

---

Usage

Running the Server

Start the MCP server locally:

python -m paper_search_mcp.server

The server runs with stdio transport, ready to accept MCP client connections.

Using with Claude Desktop

1. Install Claude Desktop: Download from Anthropic's website and install the latest version.

2. Configure MCP Server: Open or create ~/Library/Application Support/Claude/claude_desktop_config.json (Mac) or equivalent path (Windows/Linux). Add the server configuration:

   {
       "mcpServers": {
           "paper_search_server": {
               "command": "python",
               "args": [
                   "-m",
                   "paper_search_mcp.server"
               ]
           }
       }
   }
   

   Save and restart Claude Desktop.

3. Verify Integration: In Claude Desktop, look for the hammer icon (🔨) in the UI. Click it to see available tools (search_arxiv, download_arxiv, etc.).

Example Commands

• Search Papers:

  Search for "machine learning" papers on arXiv
  

  Returns a list of 10 papers (default max_results).

• Download a Paper:

  Download arXiv paper 2106.12345
  

  Saves the PDF to ./downloads (default path).

• Custom Search:

  Search PubMed for "cancer treatment" with max_results=5
  

  Returns 5 papers from PubMed.

---

Deployment

Local Deployment

For personal use or development:

nohup python -m paper_search_mcp.server &

Keeps the server running in the background. Check nohup.out for server output.

Cloud Deployment

To deploy on a cloud service (e.g., AWS EC2):

1. Set Up EC2 Instance:

   • Launch an Ubuntu instance.
   • Open port 22 (SSH) and ensure internet access.

2. Install Dependencies:

   sudo apt update
   sudo apt install python3 python3-pip git
   git clone https://github.com/yourusername/paper-search-mcp.git
   cd paper-search-mcp
   pip3 install -e .
   

3. Run Server:

   python3 -m paper_search_mcp.server
   

   Use screen or tmux to keep it running after logout:

   screen
   python3 -m paper_search_mcp.server
   # Press Ctrl+A, then D to detach
   

---

Contributing

We welcome contributions! Here's how to get started:

1. Fork the Repository: Click "Fork" on GitHub.

2. Clone and Set Up:

   git clone https://github.com/yourusername/paper-search-mcp.git
   cd paper-search-mcp
   pip install -e ".[dev]"  # Install dev dependencies (if added to pyproject.toml)
   

3. Make Changes:

   • Add new platforms in academic_platforms/.
   • Update tests in tests/.

4. Submit a Pull Request: Push changes and create a PR on GitHub.

---

Demo

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Happy researching with paper-search-mcp! If you encounter issues, open a GitHub issue.