Law Scrapper MCP

Instructions

Implementation Reference

• app.py:521-580 (handler)

  The complete handler for the 'get_act_content' tool, including registration decorator, type-annotated parameters (schema), docstring description, and execution logic that retrieves PDF URL or HTML text from the Sejm API.

  @app.tool(
      name="get_act_content",
      description="Retrieve the actual text content of a legal act in PDF or HTML format. Use for reading the full document content.",
      tags={"analysis", "content", "text", "reading"}
  )
  def get_act_text(
      publisher: Annotated[str, "Publisher code (DU for Dziennik Ustaw, MP for Monitor Polski)"],
      year: Annotated[int, "Publication year"],
      num: Annotated[Union[int, str], "Act number/position within the year"],
      format_type: Annotated[str, "Content format: 'pdf' or 'html' (default: pdf)"] = "pdf"
  ) -> str:
      """Fetches the text content of a specific legal act in PDF or HTML format.
  
      Retrieves the actual textual content of a legal act for reading and analysis.
      For PDF format, returns a download URL. For HTML format, returns the full
      HTML content of the document.
  
      Args:
          publisher: Publication code (e.g., 'DU' for Dziennik Ustaw, 'MP' for Monitor Polski).
          year: Year of publication as integer.
          num: Act number/position within the year as integer or string.
          format_type: Content format - 'pdf' or 'html'. Defaults to 'pdf'.
  
      Returns:
          str: For PDF format, returns download URL string. For HTML format, returns
               the full HTML content. Returns empty string if content not found or error.
  
      Examples:
          User asks: "Get the PDF text of DU/2020/1280":
              Parameters: publisher = 'DU', year = 2020, num = 1280, format_type = 'pdf'
              Returns: "PDF content available at: https://api.sejm.gov.pl/eli/acts/DU/2020/1280/text.pdf"
          User asks: "Get the HTML text of DU/2020/1":
              Parameters: publisher = 'DU', year = 2020, num = 1, format_type = 'html'
          User asks: "Download PDF of act MP/2023/100":
              Parameters: publisher = 'MP', year = 2023, num = 100, format_type = 'pdf'
          User asks: "Show me the HTML content of DU/2019/50":
              Parameters: publisher = 'DU', year = 2019, num = 50, format_type = 'html'
          User asks: "I need the text of act DU/2022/200 in PDF":
              Parameters: publisher = 'DU', year = 2022, num = 200, format_type = 'pdf'
      """
      logger.debug(f"get_act_text called with: publisher={publisher}, year={year}, num={num}, format_type={format_type}")
      try:
          url = f"https://api.sejm.gov.pl/eli/acts/{publisher}/{year}/{num}/text.{format_type}"
          logger.debug(f"Making GET request to: {url}")
          response = requests.get(url)
          response.raise_for_status()
  
          if format_type == "pdf":
              # For PDF, return download URL since binary content can't be displayed directly
              result = f"PDF content available at: {url}"
              logger.info(f"get_act_text retrieved PDF for act: {publisher}/{year}/{num}")
              return result
          else:
              # For HTML, return the actual content
              logger.info(f"get_act_text retrieved HTML content for act: {publisher}/{year}/{num}")
              return response.text
      except Exception as e:
          logger.error(f"Error: {e}")
          return ""

• app.py:521-525 (registration)

  The @app.tool decorator registers the get_act_text function as the 'get_act_content' MCP tool with description and tags.

  @app.tool(
      name="get_act_content",
      description="Retrieve the actual text content of a legal act in PDF or HTML format. Use for reading the full document content.",
      tags={"analysis", "content", "text", "reading"}
  )

• app.py:526-531 (schema)

  Input schema defined via type hints and Annotated metadata for publisher (str), year (int), num (Union[int,str]), format_type (str, default 'pdf'), output str.

  def get_act_text(
      publisher: Annotated[str, "Publisher code (DU for Dziennik Ustaw, MP for Monitor Polski)"],
      year: Annotated[int, "Publication year"],
      num: Annotated[Union[int, str], "Act number/position within the year"],
      format_type: Annotated[str, "Content format: 'pdf' or 'html' (default: pdf)"] = "pdf"
  ) -> str: