[Feature Request] Voice mode: add text-to-speech responses, continuous listening, and auto-submit in tap mode
Status Open
Reported on v2.1.300
Maintainer reply None cached
Activity 0 comments · opened Oct 15, 2026
Bug Description
- Voice mode only covers dictation. There is no setting to read responses aloud, no continuous listening, and no way to start voice automatically when a session begins. Recording stops after 15 s of silence or 2 min in total. - I want to work by voice end to end, with speakers connected. - autoSubmit only applies in hold mode; in tap mode I still have to press Enter to send. - Proposal: a. Opt-in setting to read the final response aloud, not tool output. b. Optional continuous or auto-start listening, with a configurable silence timeout. c. A way to dictate punctuation such as "slash" to start commands like /feedback. - Docs consulted: voice-dictation.md and settings-reference.md.
Environment Info
- Platform: darwin
- Terminal: xterm-256color
- Version: 2.1.300
- Feedback ID: 3c3977b0-f8da-45a0-affe-54836da41eb0