- Add HF_TOKEN for authenticated model downloads from Hugging Face
- Allow external faster-whisper API for transcription
- Add user info/knowledge config (name, job, title, etc.)
- Add subcommand to generate persona stub from template that creates and populates the directory for the user
- Basic UI/AppImage
- Add dynamic tag generation
- Deduplicate tags
- Investigate slow local transcription (periodically)
- Handling for GPU-based transcription (when available)
- Remove need to truncate transcript content to fit context window
- Improved TUI experience
- Unified task progress bar for all tasks
- Fix inaccurate transcription progress readout