Paper Search MCP

Instructions

Implementation Reference

• paper_search_mcp/server.py:1172-1182 (handler)

  MCP tool registration and wrapper for `download_base`.

  @mcp.tool()
  async def download_base(paper_id: str, save_path: str = "./downloads") -> str:
      """Download PDF for a paper from BASE.
  
      Args:
          paper_id: BASE paper identifier.
          save_path: Directory to save the PDF (default: './downloads').
      Returns:
          str: Path to downloaded PDF.
      """
      return base_searcher.download_pdf(paper_id, save_path)

• paper_search_mcp/academic_platforms/base_search.py:155-170 (handler)

  Actual implementation of the PDF download logic for BASE in BASESearcher class.

  def download_pdf(self, paper_id: str, save_path: str) -> str:
      """Download PDF for a BASE record.
  
      BASE often provides direct PDF links in metadata.
  
      Args:
          paper_id: BASE identifier or OAI-PMH identifier
          save_path: Directory to save PDF
  
      Returns:
          Path to saved PDF file
  
      Raises:
          NotImplementedError: If PDF cannot be downloaded
      """
      # Try parent method first (searches for PDF URL)