arXiv Search MCP Server

# Goal ✅ COMPLETED expose search function for arxiv.org papers as mcp server using the web from from: https://arxiv.org/search/advanced # Requirements ✅ ALL IMPLEMENTED * ✅ use `fastmcp` python package * ✅ search by terms, subject, date and number of result. understand the https://arxiv.org/search/advanced page for more detail * ✅ terms are manditory, other conditions are optional * ✅ try to include papers' pdf url so user can decide to download pdf or not. # Implementation Status: COMPLETE ✅ ## Files Created: - ✅ `server.py` - Main MCP server implementation with arXiv search functionality - ✅ `requirements.txt` - Python dependencies (fastmcp, requests, feedparser, urllib3) - ✅ `README.md` - Comprehensive documentation and setup guide - ✅ `test_server.py` - MCP-based test suite - ✅ `test_standalone.py` - Standalone functionality test (validated ✅) - ✅ `USAGE.md` - Detailed usage guide with examples - ✅ `mcp-config.json` - MCP client configuration example ## Features Implemented: - ✅ **Search by terms**: Required parameter, searches in titles, abstracts, and content - ✅ **Subject filtering**: Optional filter by 20 arXiv categories (physics, cs, math, etc.) - ✅ **Date range filtering**: Optional start/end date filtering in YYYY-MM-DD format - ✅ **Configurable results**: Control number of results (max 2000 per arXiv API limit) - ✅ **PDF URLs**: Direct links to paper PDFs included in every result - ✅ **Comprehensive metadata**: Authors, abstracts, categories, publication dates, DOI, etc. - ✅ **Error handling**: Robust error handling for invalid inputs and API issues - ✅ **Subject category helper**: Tool to list all available subject categories ## MCP Tools Available: 1. **`search_arxiv_papers`** - Main search function with filtering options 2. **`get_subject_categories`** - Lists available subject categories ## Testing Results: ✅ PASSED - ✅ Basic search functionality (1M+ results found) - ✅ Subject category filtering - ✅ Date range filtering - ✅ Error handling for invalid inputs - ✅ PDF URL extraction - ✅ Metadata parsing (authors, abstracts, categories, dates) ## Usage Examples: ```python # Basic search search_arxiv_papers(terms="quantum computing", max_results=10) # Advanced search with filters search_arxiv_papers( terms="machine learning", subject="cs", start_date="2023-01-01", end_date="2023-12-31", max_results=20 ) # Get available categories get_subject_categories() ``` ## Ready to Use! 🚀 The arXiv Search MCP server is fully functional and tested. Users can now: 1. Install dependencies: `pip install -r requirements.txt` 2. Run server: `python server.py` 3. Search academic papers with full metadata and PDF links