LinkedIn MCP Server

Instructions

Implementation Reference

• linkedin_mcp_server/tools/company.py:37-110 (handler)

  The main handler function that scrapes the LinkedIn company profile using the linkedin_scraper.Company class, constructs the profile data dictionary, and optionally includes employee data.

  async def get_company_profile(
      company_name: str, get_employees: bool = False
  ) -> Dict[str, Any]:
      """
      Get a specific company's LinkedIn profile.
  
      Args:
          company_name (str): LinkedIn company name (e.g., "docker", "anthropic", "microsoft")
          get_employees (bool): Whether to scrape the company's employees (slower)
  
      Returns:
          Dict[str, Any]: Structured data from the company's profile
      """
      try:
          # Construct clean LinkedIn URL from company name
          linkedin_url = f"https://www.linkedin.com/company/{company_name}/"
  
          driver = safe_get_driver()
  
          logger.info(f"Scraping company: {linkedin_url}")
          if get_employees:
              logger.info("Fetching employees may take a while...")
  
          company = Company(
              linkedin_url,
              driver=driver,
              get_employees=get_employees,
              close_on_complete=False,
          )
  
          # Convert showcase pages to structured dictionaries
          showcase_pages: List[Dict[str, Any]] = [
              {
                  "name": page.name,
                  "linkedin_url": page.linkedin_url,
                  "followers": page.followers,
              }
              for page in company.showcase_pages
          ]
  
          # Convert affiliated companies to structured dictionaries
          affiliated_companies: List[Dict[str, Any]] = [
              {
                  "name": affiliated.name,
                  "linkedin_url": affiliated.linkedin_url,
                  "followers": affiliated.followers,
              }
              for affiliated in company.affiliated_companies
          ]
  
          # Build the result dictionary
          result: Dict[str, Any] = {
              "name": company.name,
              "about_us": company.about_us,
              "website": company.website,
              "phone": company.phone,
              "headquarters": company.headquarters,
              "founded": company.founded,
              "industry": company.industry,
              "company_type": company.company_type,
              "company_size": company.company_size,
              "specialties": company.specialties,
              "showcase_pages": showcase_pages,
              "affiliated_companies": affiliated_companies,
              "headcount": company.headcount,
          }
  
          # Add employees if requested and available
          if get_employees and company.employees:
              result["employees"] = company.employees
  
          return result
      except Exception as e:
          return handle_tool_error(e, "get_company_profile")

• linkedin_mcp_server/tools/company.py:29-36 (registration)

  The @mcp.tool decorator registers the get_company_profile function with metadata like title and hints.

  @mcp.tool(
      annotations=ToolAnnotations(
          title="Get Company Profile",
          readOnlyHint=True,
          destructiveHint=False,
          openWorldHint=True,
      )
  )

• linkedin_mcp_server/server.py:29-29 (registration)

  Calls the register_company_tools function during MCP server initialization to register the tool.

  register_company_tools(mcp)

• linkedin_mcp_server/tools/company.py:37-39 (schema)

  Function signature defines the input schema: company_name (str), get_employees (bool, default False), returns Dict[str, Any].

  async def get_company_profile(
      company_name: str, get_employees: bool = False
  ) -> Dict[str, Any]:

• linkedin_mcp_server/tools/company.py:110-110 (helper)

  Error handling wrapper specific to this tool.

  return handle_tool_error(e, "get_company_profile")