YGao2005/scholar-feed-mcp

What it does

Scholar Feed MCP is an MCP server that provides semantic search capabilities over a database of 600,000+ computer science and artificial intelligence research papers. It enables researchers to search papers by meaning rather than just keywords, traverse citation networks to discover related work, extract full text content, and generate BibTeX citations for bibliographic management.

How to set up

Installation is straightforward via Node Package Manager. Run npx scholarfeedmcp init to initialize the tool. The service works anonymously out of the box, or users can optionally provide a free API key for enhanced functionality. No paid subscription is required.

Key features

• Semantic search across 600k+ CS/AI papers with meaning-based queries
• Citation graph traversal to explore related papers and research connections
• Full-text extraction and content analysis
• Embedding-based similarity search
• BibTeX export for easy reference management
• Anonymous access with optional free API key
• Integrates with Claude and other AI assistants via MCP protocol