fix (llm): improve prompts for choosing outfits

feat(combinations): improve Get Dressed LLM ranking prompt

fix(ui): restore garment images and descriptions in Ask chat

fix(ui): add bounds to rectangle geometry so detected boxes render

fix(ui): use Annotorious v3 internal annotation format for bounding boxes

feat(ui): manual garment crop with Annotorious bounding-box editor

refactor(ui): show garment detail in a side drawer

fix(ui): log dock, garment detail layout and image cache

feat(ui): garment detail panel and real-time load logs

perf(model): tune llama.cpp for cpu-basic

fix(deploy): switch to CPU inference for ZeroGPU compatibility

fix (hf): use gpu from hf space

chore (ui): Use new ui

feat(combinations): add context-aware LLM ranking for outfits

refactor(detector): replace monolithic detector with pluggable multi-backend package

feat(ui): add manual bbox annotation fallback when detection fails

refactor(detector): replace YOLOv8n with YOLOS-tiny for flat lay support

feat(ui): add capture mode selector (single/multi/video WIP)

fix(detector): lower confidence threshold to 0.15

fix(detector): download YOLO model from HuggingFace before loading

fix(catalog): ensure data persists across app restarts

refactor(vision): use detector crops for per-garment extraction

feat(detector): add YOLOv8n garment detection and cropping

feat(combinations): add outfit combination engine

feat(vision): add verbose description to garment extraction

feat(catalog): add image persistence per garment

feat(storage): add dual storage abstraction (local/S3)

feat(wardrobe): implement vision pipeline, catalog and assistant

feat (wardrobe): prepare project structure and basic gradio ui