French Tax MCP Server

Instructions

Implementation Reference

• french_tax_mcp/server.py:496-499 (registration)

  Registration of the 'get_tax_article' MCP tool using the @mcp.tool decorator with name and description.

  @mcp.tool(
      name="get_tax_article",
      description="Get information about a tax law article from legifrance.gouv.fr",
  )

• french_tax_mcp/server.py:500-526 (handler)

  Handler function registered as MCP tool 'get_tax_article'. Logs the request, calls the scraper function, and handles errors.

  async def get_tax_article_wrapper(
      article_id: str,
      ctx: Optional[Context] = None,
  ) -> Optional[Dict]:
      """Get information about a tax law article from legifrance.gouv.fr.
  
      Args:
          article_id: Article identifier (e.g., '200', '4B')
          ctx: MCP context for logging
  
      Returns:
          Dict: Dictionary containing article information
      """
      try:
          if ctx:
              await ctx.info(f"Getting tax article information for {article_id}")
  
          result = await get_tax_article(article_id)
          return result
      except Exception as e:
          if ctx:
              await ctx.error(f"Failed to get tax article: {e}")
          return {
              "status": "error",
              "message": f"Error getting tax article: {str(e)}",
          }

• french_tax_mcp/scrapers/legal_scraper.py:41-80 (helper)

  Core scraping logic in LegalScraper.get_tax_article method: constructs Legifrance URL, fetches page, parses HTML, extracts info, formats result.

  async def get_tax_article(self, article_id: str) -> Dict:
      """Scrape information about a tax law article from legifrance.gouv.fr.
  
      Args:
          article_id: Article identifier (e.g., '200', '4B')
  
      Returns:
          Dictionary containing information about the article
      """
      logger.info(f"Scraping information for article {article_id}")
  
      try:
          # Construct URL
          url = f"/codes/id/LEGITEXT000006069577/LEGIARTI000{article_id}"
  
          # Get the page
          response = await self.get_page(url)
  
          # Parse HTML
          soup = self.parse_html(response.text)
  
          # Extract article information
          article_info = self._extract_article_info(soup, article_id)
  
          return self.format_result(
              status="success",
              data=article_info,
              message=f"Successfully retrieved information for article {article_id}",
              source_url=f"{BASE_URL}{url}",
          )
  
      except Exception as e:
          logger.error(f"Error scraping article information: {e}")
          return self.format_result(
              status="error",
              message=f"Failed to retrieve article information: {str(e)}",
              data={"article": article_id},
              error=e,
          )

• french_tax_mcp/scrapers/legal_scraper.py:329-339 (helper)

  Convenience free function that delegates to the singleton LegalScraper instance's get_tax_article method.

  async def get_tax_article(article_id: str) -> Dict:
      """Scrape information about a tax law article from legifrance.gouv.fr.
  
      Args:
          article_id: Article identifier (e.g., '200', '4B')
  
      Returns:
          Dictionary containing information about the article
      """
      return await legal_scraper.get_tax_article(article_id)