## Resources

All runtime assets live under `app/resources`.

### Models (`app/resources/models`)
- `dlib/shape_predictor_68_face_landmarks.dat` (required by `GazeTrackingService`)
- `frontface_default.xml` (OpenCV face cascade)
- `haarcascade_eye.xml` (OpenCV eye cascade, optional)
- `haarcascade_smile.xml` (OpenCV smile cascade)

### Word Lists (`app/resources/word_lists`)
- `filler.txt` (comma-separated single words)
- `filler_phrase.txt` (comma-separated phrases)
- `power.txt`
- `power_phrase.txt`
- `negative.txt`
- `um_words.txt`

Notes:
- Files are comma-separated for compatibility with legacy content analysis.
- Use `analysis_utils.load_word_list("<name>.txt")` to load.