• bump version to 0.9.2
• refactor: simplify audio feedback warning UI and translations
• feat: replace deprecated ScriptProcessor with AudioWorklet API
• fix: passthrough audio issues in Extension environment
• fix: correct Int16Array to base64 conversion in GeminiClient
• fix: support locale format in PalabraAI queue status detection
• feat: add dynamic audio device switching and update documentation
• feat: add virtual speaker support for Electron environments
• refactor: remove wavtools dependency and use modern audio APIs
• docs: update documentation to reflect modern audio architecture
• fix: prevent automatic connections between virtual mic and physical devices
• docs: add CLAUDE.md project documentation for AI assistance
• fix: resolve audio initialization errors and improve stability
• fix: add missing addToPassthroughBuffer method to ModernAudioPlayer
• refactor: remove audio element pooling to fix connection issues
• fix: audio visualization and device switching issues
• refactor: simplify pulseaudio-utils to only handle virtual device creation
• perf: optimize audio chunk playback with event-driven queue and element pooling
• perf: reduce buffering and timeout intervals for faster audio streaming
• feat: remove BrowserAudioService and ElectronAudioService classes
• chore: remove fix_titles.sh script and relocate Chrome Web Store response doc
• Merge pull request #56 from kizuna-ai-lab/fix/audio-echo-issues-55
• feat(locales): add audio feedback warnings and solutions for multiple languages
• feat: add audio feedback warning translations for multiple locales
• chore(docs): remove app analytics and integration documentation