stringadmin
ec9204437d
Previously all 4 image tools generated a single image and duplicated it across 5 fixed card slots. Now: - Set (套图) mode: uses cloneSetCounts (卖点图/白底图/场景图 quantity) to determine how many images to generate. Each type gets its own createImageTask call with a type-specific prompt, so images differ by category (selling-point vs white-background vs lifestyle scene). - Preview cards are dynamically built from cloneSetCounts, not from the fixed 5-slot productSetPreviewCards template. A card labeled "卖点图 1", "卖点图 2" etc for count > 1, or just "卖点图" for count = 1. - clonePreview: shows dynamic card grid for set mode, single result for detail/model/hot modes. - setPreview: shows original image as main card, then all generated set cards in the grid. - generateSetImages: new async function that loops over each count type and generates images sequentially with distinct prompts. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>