Skip to content

Instantly share code, notes, and snippets.

@harryge00
Created May 22, 2025 11:50
Show Gist options
  • Save harryge00/683987e4e6f57f03179c5efffc8c45b2 to your computer and use it in GitHub Desktop.
Save harryge00/683987e4e6f57f03179c5efffc8c45b2 to your computer and use it in GitHub Desktop.
prompt.md

[CHARACTER=角色 (游戏 character)] [STYLE=pixel‑art] [2DTO3D=no]

今天试了一组能用ChatGPT批量创作像素角色的提示词

Create THREE cohesive sprite sheets.

─ROW 1 – 8‑View Base Model ─ • If 2DTO3D=yes → T‑pose in 8 rotations (front, 45°, side, 135°, back, 225°, other side, 315°). • If 2DTO3D=no → Natural idle pose in the same 8 rotations. • Match the chosen STYLE exactly.

─ ROW 2 – Adaptive Equipment Grid─ • Auto‑invent 8–12 distinct gear/prop items (no duplicates within or across sheets) that logically suit the CHARACTER; cover head, torso, arms, legs, feet, plus one‑two signature extras. • Render each item in its own cell at the clearest modelling angle(s), using the same STYLE. • Ensure every piece fits the ROW 1 model without clipping; if 2DTO3D=yes, favor modelling‑friendly perspectives.

─ Global Constraints ─ • Keep all three sheets visually and stylistically consistent (palette, light direction, line weight). • Use power‑of‑two canvases ≤ 2048 × 2048 px; leave padding for atlas packing. • Output only the finished sprite sheets—no explanatory text.

@harryge00
Copy link
Author

image

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment