Skip to content

Latest commit

 

History

History
59 lines (44 loc) · 2.37 KB

CLAUDE.md

File metadata and controls

59 lines (44 loc) · 2.37 KB

Claude Development Notes for Super Voice Assistant

Project Guidelines

  • Follow the roadmap and tech choices outlined in README.md

Background Process Management

  • When developing and testing changes, run the app in background using: swift build && swift run SuperVoiceAssistant with run_in_background: true
  • Keep the app running in background while the user tests functionality
  • Only kill and restart the background instance when making code changes that require a fresh build
  • Allow the user to continue using the running instance between agent sessions
  • The user prefers to keep the app running for continuous testing

Git Commit Guidelines

  • Never include Claude attribution or Co-Author information in git commits
  • Keep commit messages clean and professional without AI-related references

Completed Features

Gemini Live TTS Integration

Status: ✅ Complete and integrated into main app Key Files:

  • SharedSources/GeminiStreamingPlayer.swift - Streaming TTS playback engine
  • SharedSources/GeminiAudioCollector.swift - Audio collection and WebSocket handling
  • SharedSources/SmartSentenceSplitter.swift - Text processing for optimal speech

Features:

  • ✅ Cmd+Opt+S keyboard shortcut for reading selected text aloud
  • ✅ Sequential streaming for smooth, natural speech with minimal latency
  • ✅ Smart sentence splitting for optimal speech flow
  • ✅ 15% speed boost via TimePitch effect

Gemini Audio Transcription

Status: ✅ Complete and integrated into main app Branch: gemini-audio-feature Key Files:

  • SharedSources/GeminiAudioTranscriber.swift - Gemini API audio transcription
  • Sources/GeminiAudioRecordingManager.swift - Audio recording manager for Gemini

Features:

  • ✅ Cmd+Opt+X keyboard shortcut for Gemini audio recording and transcription
  • ✅ Cloud-based transcription using Gemini 2.5 Flash API
  • ✅ WAV audio conversion and base64 encoding
  • ✅ Silence detection and automatic filtering
  • ✅ Mutual exclusion with WhisperKit recording and screen recording
  • ✅ Transcription history integration

Keyboard Shortcuts:

  • Cmd+Opt+Z: WhisperKit audio recording (offline)
  • Cmd+Opt+X: Gemini audio recording (cloud)
  • Cmd+Opt+S: Text-to-speech with Gemini
  • Cmd+Opt+C: Screen recording with video transcription
  • Cmd+Opt+A: Show transcription history
  • Cmd+Opt+V: Paste last transcription at cursor