mcp-luma-dream-machine

Instructions

Implementation Reference

• src/luma_ai_mcp_server/server.py:401-427 (handler)

  The main handler function for the 'add_audio' tool. It validates parameters, constructs the request to the Luma API to add audio to a video generation, and returns the status.

  async def add_audio(parameters: dict) -> str:
      """Add audio to a video generation."""
      try:
          generation_id = parameters.get("generation_id")
          if not generation_id:
              raise ValueError("generation_id parameter is required")
  
          prompt = parameters.get("prompt")
          if not prompt:
              raise ValueError("prompt parameter is required")
  
          request_data = {"generation_type": "add_audio", "prompt": prompt}
          if "negative_prompt" in parameters:
              request_data["negative_prompt"] = parameters["negative_prompt"]
  
          result = await _make_luma_request(
              "POST", f"/generations/{generation_id}/audio", request_data
          )
  
          return (
              f"Audio generation initiated for generation {generation_id}\n"
              f"Status: {result['state']}\n"
              f"Prompt: {prompt}"
          )
      except Exception as e:
          logger.error(f"Error in add_audio: {str(e)}", exc_info=True)
          return f"Error adding audio to generation {generation_id}: {str(e)}"

• src/luma_ai_mcp_server/server.py:154-159 (schema)

  Pydantic input schema for the 'add_audio' tool, defining required generation_id and prompt, with optional negative_prompt and callback_url.

  class AddAudioInput(BaseModel):
      generation_id: str
      prompt: str
      negative_prompt: Optional[str] = None
      callback_url: Optional[str] = None

• src/luma_ai_mcp_server/server.py:528-532 (registration)

  Tool registration in the MCP server's list_tools() function, specifying name, description, and input schema for 'add_audio'.

  Tool(
      name=LumaTools.ADD_AUDIO,
      description="Adds audio to a video generation",
      inputSchema=AddAudioInput.model_json_schema(),
  ),

• src/luma_ai_mcp_server/server.py:579-581 (registration)

  Dispatcher in call_tool() that routes calls to the 'add_audio' handler function.

  case LumaTools.ADD_AUDIO:
      result = await add_audio(arguments)
      return [TextContent(type="text", text=result)]

• src/luma_ai_mcp_server/server.py:101-101 (helper)

  Enum value in LumaTools defining the tool name 'add_audio'.

  ADD_AUDIO = "add_audio"