MCP Ahrefs

Instructions

Args:
    min_value: Minimum value (default: 1)
    max_value: Maximum value (default: 100)
    
Returns:
    Dictionary containing the random number and range info

Implementation Reference

• mcp_ahrefs/tools/example_tools.py:40-58 (handler)

  The main tool handler function that generates a random integer between min_value and max_value, returning a dictionary with the number, range string, and timestamp. Includes input validation.

  async def random_number(min_value: int = 1, max_value: int = 100) -> Dict[str, Any]:
      """Generate a random number within a specified range.
      
      Args:
          min_value: Minimum value (default: 1)
          max_value: Maximum value (default: 100)
          
      Returns:
          Dictionary containing the random number and range info
      """
      if min_value > max_value:
          raise ValueError("min_value must be less than or equal to max_value")
      
      number = random.randint(min_value, max_value)
      return {
          "number": number,
          "range": f"{min_value}-{max_value}",
          "timestamp": time.time()
      }

• mcp_ahrefs/tools/example_tools.py:178-183 (registration)

  List of regular example tools including the random_number handler, which gets imported and looped over for registration in mcp_ahrefs/server/app.py.

  example_tools = [
      echo_tool,
      get_time,
      random_number,
      calculate_fibonacci
  ]

• mcp_ahrefs/server/app.py:98-112 (registration)

  Registration loop that applies SAAGA decorators (exception_handler and tool_logger) to each tool in example_tools (including random_number) and registers them to the MCP server using mcp_server.tool(name=tool_name).

  for tool_func in example_tools:
      # Apply SAAGA decorator chain: exception_handler → tool_logger
      decorated_func = exception_handler(tool_logger(tool_func, config.__dict__))
      
      # Extract metadata from the original function
      tool_name = tool_func.__name__
      
      # Register the decorated function directly with MCP
      # This preserves the function signature for parameter introspection
      mcp_server.tool(
          name=tool_name
      )(decorated_func)
      
      unified_logger.info(f"Registered tool: {tool_name}")