Continue the previous task of cherry-picking TTS/STT code from `c-agent-embedding` to `vpcp-embedding`.

**Progress so far:**
- Created `src/utils/transcriptApi.ts` file
- Added import for transcriptApi to Bot.tsx

**Remaining work:**
1. Add `isVoiceChat` signal to Bot.tsx
2. Add TTS state signals (ttsEnabled, ttsError, etc.) and TTS-related variables (ttsWS, audioContext, etc.)
3. Add TTS functions: ttsConnect, ttsDisconnect, ttsStartNewStream, ttsFinalizeStream, ttsAddTokenToLine, ttsSendText
4. Add STT functions: handleTranscriptAudio, handleRecordingToText
5. Modify onRecordingStopped to use transcription and voice chat mode
6. Add voice chat UI toggle button in the UI
7. Run lint-fix and type-check to ensure code quality

**Definition of Done:**
- All TTS and STT code from `c-agent-embedding` is successfully integrated into `vpcp-embedding` branch
- The code compiles without errors
- Run lint-fix and type-check to ensure code quality

Do not create or modify any markdown document files or code examples.