feat: Voice Suite — TTS v1, Assistant mode, Language Tutor#19
Merged
Conversation
Word-level alignment with Levenshtein tolerance + Cyrillic-aware normalization. Recovered from a Serena-MCP cross-write during a parallel Gemini delegation run; 5/5 unit tests green. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- speak/synthesize gain a `rate` param (WinRT SpeakingRate, clamped 0.5–6.0) - pick_voice_for_text: Cyrillic≥30% → ru voice, else en (auto when no explicit voice); 3 unit tests - AppSettings: tts_enabled / tts_voice_id / tts_rate; agent-bridge sink now respects them (only speaks when enabled, uses chosen voice + rate) - new Voice settings card (toggle, voice picker, rate slider, test button), Sidebar entry, en/ru i18n - tts_speak command + bindings gain optional rate Verified: cargo check + cargo test --lib tts (3/3 voice tests) + tsc + eslint, all green. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…provider Reuses the configured post-processing OpenAI-compatible provider/model/key (llm_client::send_chat_completion_with_schema) with an Echo voice-assistant system prompt. New assistant module + assistant_ask command, settings (assistant_enabled / assistant_system_prompt), and an Assistant settings card (toggle, system-prompt editor, try-it box with speak-reply via TTS). en/ru i18n. Stacked on feat/tts-v1. Verified: cargo check + cargo test --lib assistant + tsc + eslint green. Built by Gemini CLI (Serena disabled) + integration fixes. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…g UI Tutor settings card (enable toggle, reference + spoken inputs, Score button) rendering the ScoreReport: overall %, note, and per-word matched/unmatched chips. Wires the tutor_score command + ScoreReport/WordScore types into bindings; Sidebar entry + en/ru i18n. Completes feat/tutor (backend was e46b661). Built by Gemini CLI (Serena disabled) + integration fixes (Result-shape, types). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
# Conflicts: # src-tauri/src/lib.rs # src-tauri/src/settings.rs # src/bindings.ts # src/components/Sidebar.tsx # src/i18n/locales/en/translation.json # src/i18n/locales/ru/translation.json
check:translations (CI code-quality) requires all 19 languages to carry every en key. TTS/Assistant/Tutor added 28 new keys (en+ru localized); the other 17 locales now carry English fallback values (i18next falls back to en at runtime anyway). Includes the one-shot fill script used to generate them. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
One stack delivering three voice/AI features toward "Echo > Wispr Flow" — a speech engine, a personal assistant, and a language tutor.
feat/tts-v1): Echo speaks. WinRTSpeakingRate(clamped 0.5–6.0), language-aware voice auto-selection (Cyrillic ≥30% →ruvoice, elseen), settings (tts_enabled/tts_voice_id/tts_rate), a Voice settings card (toggle · voice picker · rate slider · test button). The Agent Bridge sink now respects these settings.feat/assistant-mode): a personal assistant that round-trips through the already-configured post-processing OpenAI-compatible provider (llm_client::send_chat_completion_with_schema) with an Echo voice-assistant system prompt.assistant_askcommand, settings (assistant_enabled/assistant_system_prompt), and an Assistant card (toggle · system-prompt editor · try-it box · speak-reply via TTS).feat/tutor): text-level pronunciation/fluency scoring. Word-level alignment with Levenshtein tolerance + Cyrillic-aware normalization →ScoreReport(overall %, per-word matched/unmatched, feedback note).tutor_scorecommand,tutor_enabledsetting, and a Tutor practice card.All three add
en+rui18n and Sidebar entries.Test Plan
cargo test --lib— 164 passed, 0 failed (incl.tts::voice_pick_tests×3,assistant::tests,tutor::tests×5)npx tsc --noEmit— cleannpm run lint(eslint + i18n) — cleannpm run format(prettier + cargo fmt) — clean🤖 Generated with Claude Code