Scholar Search MCP

A MCP server that integrates the Semantic Scholar API and arXiv API so AI assistants (e.g. Claude, Cursor) can search and fetch academic paper metadata.

Features

• Search papers – Keyword search with parallel merge from Semantic Scholar and arXiv; optional year and venue filters (venue applies to Semantic Scholar only)
• Paper details – Full metadata (title, authors, abstract, citations, etc.)
• Citations & references – Papers that cite or are cited by a given paper
• Author info – Author profile and paper list
• Batch lookup – Fetch up to 500 papers in one call
• Recommendations – Similar papers for a given paper
• arXiv LaTeX source – Download and extract the source tarball from https://arxiv.org/src/{id} (tool: download_arxiv_source)

Installation

pip install scholar-search-mcp

Configuration

Claude Desktop

Edit the config file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Add:

{
  "mcpServers": {
    "scholar-search": {
      "command": "python",
      "args": ["-m", "scholar_search_mcp"],
      "env": {
        "SCHOLAR_SEARCH_ENABLE_SEMANTIC_SCHOLAR": "true", // enable https://www.semanticscholar.org/
        "SCHOLAR_SEARCH_ENABLE_ARXIV": "true" // enable https://arxiv.org/
      }
    }
  }
}

If you have API keys (optional but recommended for search):

{
  "mcpServers": {
    "scholar-search": {
      "command": "python",
      "args": ["-m", "scholar_search_mcp"],
      "env": {
        "SEMANTIC_SCHOLAR_API_KEY": "your-semantic-scholar-api-key-here",
        "SCHOLAR_SEARCH_ENABLE_SEMANTIC_SCHOLAR": "true", // enable https://www.semanticscholar.org/
        "SCHOLAR_SEARCH_ENABLE_ARXIV": "true" // enable https://arxiv.org/
      }
    }
  }
}

Cursor

Add an MCP server in Cursor settings with the same command, args, and env as above.

API keys (optional)

search_papers queries enabled sources in parallel and merges results by title:

1. Semantic Scholar – Works without a key with lower limits. Set SEMANTIC_SCHOLAR_API_KEY for higher limits.
2. arXiv – No key required.

Enable/disable search channels

Control which sources are used in search_papers via environment variables (default: all enabled):

Example: CORE and arXiv only (skip Semantic Scholar):

"env": {
  "SCHOLAR_SEARCH_ENABLE_SEMANTIC_SCHOLAR": "false"
}

arXiv source downloads (download_arxiv_source)

Optional default parent directory for extracted sources when the tool is called without output_dir:

Tools

Testing with MCP Inspector

npm install -g @modelcontextprotocol/inspector
mcp-inspector python -m scholar_search_mcp

License

MIT

Links

• Semantic Scholar API
• arXiv API User's Manual