Paper Search MCP

Instructions

Implementation Reference

• paper_search_mcp/server.py:538-547 (handler)

  The MCP tool wrapper function 'read_pubmed_paper' which delegates the request to the 'pubmed_searcher' instance.

  async def read_pubmed_paper(paper_id: str, save_path: str = "./downloads") -> str:
      """Read and extract text content from a PubMed paper.
  
      Args:
          paper_id: PubMed ID (PMID).
          save_path: Directory where the PDF would be saved (unused).
      Returns:
          str: Message indicating that direct paper reading is not supported.
      """
      return pubmed_searcher.read_paper(paper_id, save_path)

• paper_search_mcp/academic_platforms/pubmed.py:111-124 (handler)

  The actual implementation of 'read_paper' within the 'PubMedSearcher' class, which returns a message explaining that reading full text is not supported.

  def read_paper(self, paper_id: str, save_path: str = "./downloads") -> str:
      """Attempt to read and extract text from a PubMed paper.
  
      Args:
          paper_id: PubMed ID (PMID)
          save_path: Directory for potential PDF storage (unused)
  
      Returns:
          str: Error message indicating PDF reading is not supported
      """
      message = ("PubMed papers cannot be read directly through this tool. "
                "Only metadata and abstracts are available through PubMed's API. "
                "Please use the paper's DOI or URL to access the full text on the publisher's website.")
      return message