mpc-sintific-search-by-ai-for-ai

mpc-sintific-search-by-ai-for-ai

0
research_and_data

This project is a Model Control Protocol implementation designed to provide academic search capabilities for Cursor, integrating multiple search engines like Google Scholar and PubMed. It features Tor proxy usage to bypass rate limits, enhancing research efficiency via scientific search tools.

Scientific Search MCP for Cursor

This project implements the Model Control Protocol (MCP) for Cursor to provide scientific search capabilities across various academic search engines.

Features

  • Integration with Cursor using the MCP protocol
  • Access to multiple search engines:
    • DuckDuckGo for general web searches
    • Google Scholar for academic papers
    • PubMed for biomedical literature
    • arXiv for fields like physics and computer science
    • Semantic Scholar via Tor for AI-powered research

Installation

  1. Clone the repository and navigate into it.
  2. Install required dependencies via pip. Additional dependencies may be needed for specific engines.

Usage

  1. Start the MCP server with the appropriate script.
  2. Connect to the MCP server in Cursor.
  3. Use the scientific search tools to search various databases.

Notes

  • Proxy usage helps avoid rate limits for services like Google Scholar and Semantic Scholar.
  • For PubMed, set the PUBMED_EMAIL environment variable for email tracking.

Troubleshooting

  • Ensure Tor is running for Semantic Scholar searches if connection issues occur.
  • Expect slower responses due to Tor routing.