mirror of
https://github.com/nesquena/hermes-webui.git
synced 2026-05-26 11:40:26 +00:00
28a7b78ae4
Adds Web Speech API based text-to-speech for agent responses. Per-message speaker button in the action row, auto-read toggle for last assistant response, voice/rate/pitch controls in Settings. All TTS preferences stored in localStorage. Opt-in: TTS hidden by default until enabled in Settings. Strips markdown, code blocks, and MEDIA: paths via _stripForTTS() before speaking. Pauses synthesis when composer is focused. 26 regression tests in tests/test_499_tts_playback.py. i18n keys added for all 7 locales (en, ru, es, de, zh, zh-Hant, pt, ko). Closes #499. Co-authored-by: Felipe Linhares <fecolinhares@users.noreply.github.com>