Give Claude Code a voice with multi-engine text-to-speech. Speaks responses aloud using 5 TTS engines (edge-tts, ElevenLabs, Kokoro, macOS say, espeak-ng). Free by default. edge-tts uses Microsoft Neural voices with no API key. Includes a Stop hook for automatic speech, a /speak skill for on-demand use, and an MCP server with speak, set_engine, get_status, and list_voices tools. Text is automatically filtered to remove code blocks, API keys, file paths, and markdown before speaking.