Multi Fetch MCP Server

Instructions

Implementation Reference

• src/mcp_server_multi_fetch/server.py:357-376 (handler)

  The 'search' tool handler within call_tool, which uses the firecrawl_client to perform web searches.

  if name == "search":
      try:
          args = Search(**arguments)
      except ValueError as e:
          raise McpError(ErrorData(code=INVALID_PARAMS, message=str(e)))
      try:
          if firecrawl_client is None:
              raise McpError(ErrorData(code=INTERNAL_ERROR, message="Firecrawl client is not initialised"))
          # Firecrawl v2: search(query, options?) with limit and scrapeOptions
          result = await firecrawl_client.search(
              args.query,
              options={"limit": args.limit, "scrapeOptions": {"formats": ["markdown", "links"]}},
          )
      except Exception as e:
          raise McpError(ErrorData(code=INTERNAL_ERROR, message=f"Failed to search via Firecrawl SDK: {e!r}"))
      try:
          json_text = result.model_dump_json()
      except AttributeError:
          json_text = json.dumps(result)
      return [TextContent(type="text", text=json_text)]

• src/mcp_server_multi_fetch/server.py:189-193 (schema)

  Pydantic model defining the input schema for the 'search' tool.

  class Search(BaseModel):
      """Parameters for searching using Firecrawl search API."""
      query: Annotated[str, Field(description="Search query string")]
      limit: Annotated[int, Field(default=10, description="Maximum number of results to return.", ge=1)]

• src/mcp_server_multi_fetch/server.py:247-252 (registration)

  Registration of the 'search' tool in the list_tools function.

  Tool(
      name="search",
      description="""Searches the web using the Firecrawl search API and scrapes results in markdown and link formats by default.""",
      inputSchema=Search.model_json_schema(),
  ),
  ]