Reusable Image Style Template Builder
You get a reusable style template, with the look locked in and the scene left as variables, plus a finished image prompt. Fixing the style traits separately from the content keeps results consistent across many images.
The prompt
Act as an art director who turns visual references into reusable image-generation prompts.
Task: Turn the style description below into a fixed style lock plus a fill-in scene, then produce one final image prompt for the subject I give you.
Style description: {style_description}
Subject and scene: {scene_details}
Aspect ratio and headline text: {format_and_text}
Steps:
1. Extract the 5 to 7 visual traits that define the style (palette, lighting, composition, typography, texture, mood) and write them as a short "style lock".
2. List what must be avoided so the result does not copy any original subject, face or wording.
3. Build a template with the style lock fixed and the scene, aspect ratio and text as variables.
4. Fill the variables with my details and write the final prompt as one paragraph, stating any headline text exactly in quotation marks.
Output:
Four labelled sections: Style lock, Avoid list, Template (variables in capitals), Final prompt. No commentary beyond these.
Change the parts in curly braces. Use it when you need a set of images, such as posters or ads, that share one look but show different subjects.
What happened when we ran it
We sent
Act as an art director who turns visual references into reusable image-generation prompts. Task: Turn the style description below into a fixed style lock plus a fill-in scene, then produce one final image prompt for the subject I give you. Style description: A monochrome noir editorial poster. Black-and-white photographic portrait with a deep charcoal background and high contrast. A giant lowercase headline sits on the left in three short lines, with the first word in a white label. Lots of empty space, a tight close crop and shallow depth of field give it a serious, late-night mood. Subject and scene: A tired architect with silver hair studies a folded blueprint in a dim concrete studio after midnight. She holds a rolled plan tube and a pencil low in one hand. The background is a soft charcoal wall and the blurred edge of a drafting table. Aspect ratio and headline text: 16:9 landscape, subject weighted to the right. Headline: "focus / outlasts / noise." Small secondary text: "studio log 02:14". Steps: 1. Extract the 5 to 7 visual traits that define the style (palette, lighting, composition, typography, texture, mood) and write them as a short "style lock". 2. List what must be avoided so the result does not copy any original subject, face or wording. 3. Build a template with the style lock fixed and the scene, aspect ratio and text as variables. 4. Fill the variables with my details and write the final prompt as one paragraph, stating any headline text exactly in quotation marks. Output: Four labelled sections: Style lock, Avoid list, Template (variables in capitals), Final prompt. No commentary beyond these.
The model replied
Run on Claude Sonnet 5.5, 02 Nov 2026. Automatic check: 4 out of 5.
Mentioned for
ChatGPT, Claude, Nano Banana Pro
Where it came from
Found on GitHub, shared by VigoZhao: VigoZhao/AI-Visual-Prompt-Cookbook.
This is our own version of the idea, written and tested by us. The original may differ.