v1.5.1

# FluidVoice v1.5.1

## Audio Device Improvements

- Added new "Sync with System Settings" option for audio devices
  When enabled, FluidVoice and macOS share the same audio devices (bidirectional sync)
  When disabled, FluidVoice uses independent audio device selection without affecting system settings
- Fixed audio device binding issues when using external microphones
- Fixed issue where FluidVoice would hijack system audio routing when connecting headphones

## Bluetooth Headset Support

- Added proper audio engine reset to fully release Bluetooth microphone connection

## New Features
- FluidVoice gets a new menubar icon refresh and a new Logo!
- Added Intel (x86_64) architecture support -- universal binary now works on both Apple Silicon and Intel Macs

- Added support for displaying AI "thinking tokens" (model reasoning) in Command Mode and Write Mode
- Added toggle in Settings to enable/disable debug logs
- Added toggle to enable/disable saving transcription history 
- Added enable/disable toggle for Command Mode and Write Mode shortcuts
- Added additional Whisper model options for more flexibility in transcription quality vs speed

## Stability Improvements

- Fixed crash-on-launch issue related to AttributeGraph race condition
- Fixed race condition during longer transcription sessions on some devices
- Fixed JSON fallback escaping in TerminalService for more reliable command execution
- Fixed issue where the audio input device picker could sometimes appear empty
- Fixed toggle for "Enable filler words removal" not updating properly
- Refactored ContentView for better stability and maintainability

## Architecture Changes

- Centralized service initialization to prevent startup crashes