Worked on the openclaw/openclaw repository to standardize cross-channel text-to-speech playback by implementing Opus-based audio delivery with voice bubble support for Feishu, WhatsApp, and Telegram. This feature unified media formats across major messaging platforms, reducing the need for channel-specific customization and maintenance. Leveraging TypeScript for backend and full stack development, the work focused on API integration and audio processing to ensure consistent playback and improved user experience. By enabling Opus audio and voice bubbles where only MP3 was previously supported, the solution minimized playback failures and streamlined the handling of TTS media across multiple communication channels.