Enables seamless text-to-speech and speech-to-text capabilities using local and cloud-based providers.
This skill empowers Claude with a voice interface, allowing you to dictate instructions and listen to responses directly within your terminal environment. By integrating local tools like macOS 'say' and Apple Silicon-optimized mlx-whisper alongside high-fidelity cloud services like ElevenLabs, it provides a flexible solution for hands-free coding. Whether you need to transcribe meeting notes, listen to long documentation, or interact with Claude through a conversational loop, this skill handles the complexity of audio processing and hardware integration.
Key Features
011 GitHub stars
02Bi-directional voice interaction with the interactive /voice command loop
03Provider validation tools to ensure microphone and API configurations are correct
04Local STT optimized for Apple Silicon using mlx-whisper
05High-fidelity cloud TTS and STT via ElevenLabs API integration
06Support for local macOS system voices with adjustable rates and output formats
Use Cases
01Dictating complex refactoring instructions to Claude while away from the keyboard
02Transcribing local audio files or meeting recordings directly into your workspace
03Auditory monitoring of build statuses or long-running script outputs
