怎麼填變量：[一個 Z 世代的工作坊女孩] 換成你的角色定位（"一個熱愛登山的大學男生""一個咖啡店店長"），外貌、服裝和包裏的物品清單也要一起改成符合人設的東西——物品清單寫得越具體，角色越有故事感。[凌亂的栗棕色] 是髮色和髮質。

示例圖兩張：第一張左側是扎紅髮圈高馬尾、栗棕頭髮、掛着口哨掛繩的女孩大胸像，右側是正面和背面全身——奶白紅條衛衣、黑色束腳褲、舊運動鞋、單肩托特包裏插着路錐、本子和彩色布料；第二張是作者延伸出的四面轉身圖（正、側、背、側）。

常見問題：

• 三個形象不一致（髮型、包變了）：強調"三者完全一致"，並把關鍵特徵（紅髮圈、口哨）寫在角色段落裏。
• 數量清單出錯：數量約束可以適當放寬，重點保留"服裝單品和顏色"。
• 後續出圖保持一致：把這張設定表作為參考圖上載，再寫"按參考圖裏的角色生成……"。

適合：短影片 / 動畫角色設定、遊戲 NPC 設計、漫畫人設、AI 連續出圖的角色參考。

英文原版

text
Goal: Create an anime character design sheet for {argument name="character name" default="a Gen Z app-workshop girl"}, showing a practical, stylish young woman as a reusable visual asset for video scenes.

Canvas: Wide horizontal character sheet, 3:2 aspect ratio, light warm gray background, subtle paper texture, clean vertical divider between the portrait area and the full-body turnaround area.

Layout: Use exactly 3 character depictions: 1 large bust portrait on the left taking about 40% of the canvas, 1 full-body front view in the center-right, and 1 full-body back view on the far right. Keep all three consistent in face, hair, outfit, proportions, and accessories.

Character details: Young woman in her early twenties with warm tan skin, large expressive brown eyes, soft confident smile, slim athletic build, delicate nose, natural brows, and small hoop earrings. She has {argument name="hair color" default="messy chestnut brown"} hair tied in a high ponytail with a red scrunchie, loose strands and flyaways framing the face, voluminous layered texture, and energetic movement.

Outfit: Casual urban creative-worker clothing with exactly 6 main outfit pieces: 1 oversized cream sweatshirt with muted red collar, cuffs, and sleeve stripes; 1 pair of loose black jogger pants with drawstring waist and gathered ankles; 1 pair of worn off-white sneakers with red-brown accents; 1 black lanyard necklace holding a small rectangular device or badge; 1 pair of small silver hoop earrings; 1 red hair scrunchie.

Bag and visible contents: Include exactly 1 large off-white canvas tote bag carried over one shoulder in the full-body views. The tote is stuffed with exactly 7 visible items: 1 orange traffic cone, 1 blue notebook or tablet, 1 dark rectangular device, 1 rolled orange fabric or scarf, 1 mustard yellow cloth, 1 teal cloth, and 1 reddish bundled item. In the front view, the tote hangs at her side; in the back view, the tote rests against her back/hip with the contents still visible.

Visual style: Detailed hand-drawn anime concept art, clean ink outlines, expressive linework, soft cel shading, muted earthy colors, slight sketch texture, subtle hatching on fabric folds, natural character-sheet lighting, no dramatic shadows. Make the pose relaxed and approachable, with the front full-body figure standing casually and the back view showing the same outfit and ponytail clearly.

Constraints: No text labels, no watermark, no extra props beyond the listed outfit and bag contents, no additional characters, maintain accurate turnaround consistency, keep the background plain and unobtrusive.

改編自 Guillaume Aubry 發佈、YouMind-OpenLab/awesome-gpt-image-2 收錄的提示詞，許可證 CC BY 4.0，本站已翻譯並改寫（修改內容見 changes）；示例圖來自同一條目。