Speech engine, using Qwen TTS model

Key features

• speak/cheer model tool (log-only voice/event)
• Daily greeting schedulerUse agent.followup to let the Agent generate "Welcome + Interesting Stories/News" by itself (or configure a fixed copy to read it directly)
• Tone configuration driverconfig.styles (timble directory) + config.defaultStyle
• No built-in UI (no icons/buttons are rendered) - only low-level behaviors are provided: automatic reading, audio playback, cheer cards, and the configuration/selection interface is completely left to the consumer plug-in to implement
• Audio queue playback (fetch WAV →, sequential playback without overlap)