senseaudio-voice

What This Skill Does

SenseAudio Voice is a high-performance, dual-engine voice interaction skill designed for OpenClaw. It integrates both Text-to-Speech (TTS) and Automatic Speech Recognition (ASR) to provide a seamless natural language processing experience. By intelligently routing requests between SenseAudio (for high-quality Chinese voice generation) and Microsoft Edge TTS (for robust international language support in English and Japanese), this skill ensures optimal audio quality regardless of the user's target language. It features automated language detection, a flexible voice selection system, and a robust file management structure that organizes output audio by date.

Installation

To install this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/cowboy231/senseaudio-voice-cn Ensure you have the requests Python library installed. For users targeting Chinese audio, obtain an API key from the SenseAudio platform and save it to your ~/.openclaw/openclaw.json under env.SENSE_API_KEY.

Use Cases

• Educational Tools: Perfect for children's learning scenarios, utilizing the child_0001_a voice to provide interactive feedback or homework reminders.
• Multilingual Assistant: Acts as a bridge for AI agents that need to communicate effectively in Chinese, English, or Japanese without switching backend configurations manually.
• Accessibility Services: Provides a voice output layer for automation scripts, making AI agents more interactive and approachable for non-technical users.
• Content Creation: Quickly batch-generate voice-over files for multimedia projects, saved with intuitive timestamped naming conventions.

Example Prompts

1. "OpenClaw, please read this message aloud using the standard Chinese child voice: 'It is time to start your math practice now.'"
2. "Convert the following text to an English audio file: 'Welcome to the system, how can I assist you today?' and play it back for me."
3. "Summarize the latest weather report and narrate it using a professional male voice, then save the output to my workspace."

Tips & Limitations

• Engine Selection: While the auto mode is excellent for most, you can force specific engines using --engine senseaudio or --engine edge if you prefer a specific sound profile.
• Storage: Audio files are stored in {workspace}/audio/YYYY-MM-DD/. Regularly clear this directory if you generate large volumes of audio to maintain workspace cleanliness.
• Dependencies: This skill is optimized for simplicity; it relies primarily on the requests library and avoids heavy local decoding dependencies, making it fast and portable. Note that the Chinese engine requires a verified mainland mobile number, whereas the Edge TTS engine is globally accessible without authentication.