01Adaptive Taste Engine: Auto-learns user preferences across seven distinct dimensions including tone and density.
02Taste Calibration: Periodically validates its understanding of your creative style to prevent drift.
03Controlled Divergence Spectrum: Generates options across Safe, Stretch, and Wild risk profiles.
04Iterative Feedback Loop: Systematically refines outputs based on explicit and implicit feedback signals.
05Preference Memory: Persists learned styles and risk levels in a dedicated preferences file for consistent performance.
063 GitHub stars