Consolidate soundboard into main (natural TTS voice) #6

Merged
charles merged 4 commits from consolidate-main into main 2026-11-05 19:01:07 +00:00
Owner

Merges soundboard (including #5, the natural Kokoro TTS voice) into main, which already matches desktop-soundboard. main becomes the only working branch, so image builds now trigger on pushes to main.

🤖 Generated with Claude Code

Merges `soundboard` (including #5, the natural Kokoro TTS voice) into `main`, which already matches `desktop-soundboard`. `main` becomes the only working branch, so image builds now trigger on pushes to `main`. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
Adds an on-device neural text-to-speech voice (Kokoro-82M) as the
default for English read-aloud, with the system voice as fallback.

- Desktop: the main process runs the fp32 model on onnxruntime-node
  (~0.35 s for a 4 s sentence), downloaded once to userData and
  SHA-256 verified. The native runtime is unpacked from asar and
  trimmed to the target platform.
- Web: a WebAssembly worker runs the q8 model from Cache Storage.
- Accessibility settings gain a "Natural voice" toggle and a voice
  picker (12 US/UK voices); English locales only for now.
- CSP allows the Hugging Face model hosts.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Merge pull request 'feat(voice): natural Kokoro TTS voice for read-aloud' (#5) from feature/kokoro-tts into soundboard
All checks were successful
build images / Build and push images (push) Successful in 12m19s
604e646787
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Sign in to join this conversation.
No reviewers
No labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
charles/fluxer!6
No description provided.