Prompt: Regenerate this image with Moe-chan as the subject — a freshly re-rendered sibling of the source (same shot, not a traced copy or paste-over). Award-winning Pixiv-tier anime illustration.

Moe-chan (keep recognizable): round soft face, fair skin with small cheek blush; pale mint / seafoam hair in TWO top buns (odango) with blunt straight bangs, short and shoulder-length; large bright orange / amber eyes (never blue, green, or pink); petite youthful-teen proportions, not chibi. Always include at least one orange accent and one orange / yellow star — off the head by default (earring, brooch, pendant, belt charm, fabric print); on the head only if the scene calls for hair accessories.

Expression follows the source's mood — cheerful by default, contemplative if quiet, soft if intimate, focused if tense; eye color stays constant.

Outfit: invent one that fits the scene's fashion register; do not reuse a fixed costume.

Preserve the shot: same camera viewpoint, framing, and pose family (sitting stays sitting, standing stays standing, hand-near-face stays hand-near-face), with natural seed-like variation. Keep the source's location, lighting, time of day, mood, key props, and color palette. If the source is candid (unposed, gaze away), Moe-chan stays candid — no peace sign, no smile-for-camera. Match the source's art style: line weight, shading, color saturation, and finish.

If notes are provided, incorporate them while keeping Moe-chan's identity.

Output: clean anatomy and hands, no text, no watermark.

Maker her hapy Model: GPT Image 2.5 Sunburst v2.5