MusicGPT MCP Server

Instructions

Implementation Reference

• src/index.ts:900-919 (handler)

  The handler function that implements the core logic for the voice_changer tool. It validates inputs, makes a POST request to the MusicGPT API's /voicetovoice endpoint, and returns the task details with instructions for status checking.

  private async handleVoiceChanger(args: any) {
    if (!args.audio_url || !args.voice_id) {
      throw new McpError(ErrorCode.InvalidParams, "audio_url and voice_id are required");
    }
  
    const response = await this.axiosInstance.post("/voicetovoice", {
      audio_url: args.audio_url,
      voice_id: args.voice_id,
      webhook_url: args.webhook_url,
    });
  
    return {
      content: [
        {
          type: "text",
          text: `Voice conversion started!\n\n${JSON.stringify(response.data, null, 2)}\n\nUse get_conversion_by_id with the task_id to check the status.`,
        },
      ],
    };
  }

• src/index.ts:211-226 (schema)

  Input schema defining the parameters for the voice_changer tool: audio_url (required), voice_id (required), and optional webhook_url.

  type: "object" as const,
  properties: {
    audio_url: {
      type: "string",
      description: "URL of the audio file to convert",
    },
    voice_id: {
      type: "string",
      description: "Target voice model ID (use get_all_voices to find IDs)",
    },
    webhook_url: {
      type: "string",
      description: "URL for callback upon completion",
    },
  },
  required: ["audio_url", "voice_id"],

• src/index.ts:207-228 (registration)

  Tool registration entry in the TOOLS array, which is returned by the listTools handler. Includes name, description, and input schema.

  {
    name: "voice_changer",
    description: "Convert audio from one voice to another using AI voice models",
    inputSchema: {
      type: "object" as const,
      properties: {
        audio_url: {
          type: "string",
          description: "URL of the audio file to convert",
        },
        voice_id: {
          type: "string",
          description: "Target voice model ID (use get_all_voices to find IDs)",
        },
        webhook_url: {
          type: "string",
          description: "URL for callback upon completion",
        },
      },
      required: ["audio_url", "voice_id"],
    },
  },

• src/index.ts:677-678 (registration)

  Switch case in the CallToolRequestSchema handler that routes execution to the specific handleVoiceChanger method.

  case "voice_changer":
    return await this.handleVoiceChanger(args);