[Feature]: add voice activation (wake word) support for hands-free workspace interaction

[Albert-Weasker]

Problem

Holaboss currently relies on keyboard / UI-driven interaction. This makes basic usage harder for non-technical users and adds friction for common workflows (e.g. triggering tasks, switching context).

Proposal

Add voice activation (wake word) support so users can trigger actions hands-free.

A reasonable first step would be:

• introduce a lightweight wake word listener (local, low CPU)
• add a speech-to-text (STT) layer and convert voice input into standard runtime requests
• reuse the existing request pipeline so voice input behaves the same as text input
• expose basic controls in the desktop UI (enable/disable voice, push-to-talk fallback)

Keep the implementation modular so STT can be swapped (local vs provider-based).

Alternatives considered

Keep interaction text-only and rely on UI improvements, but that doesn’t address hands-free usage or reduce friction for non-technical users.

Another option is push-to-talk only (no wake word), which is simpler but less seamless and doesn’t enable ambient / continuous interaction.

Impact

This mainly affects the desktop layer.

[jeffreyliimerch]

Thanks for writing this up so clearly. Voice activation and wake-word driven interaction would be a strong addition to the desktop experience, especially for reducing friction in everyday use.

This is in a direction we already want to invest in, and voice is on our roadmap. We do not have a concrete timeline to share yet, but it is something we plan to explore further. Stay tuned for future updates.

[621625]

I like this idea as well. When LLM's happened people got obsessed with them. Somewhat unfortunately. We should focus on voice because it is really the natural way that humans communicate with each other, not through their fingertips. At least not on the first date.

#734598