A text-to-speech (TTS) MCP server based on FishSpeech, supporting voice reference saving and can be integrated with MCP-compatible LLMs such as Dive.