All products

Write an image prompt from a reference with AI

Give your agent an image you can adapt and a fresh topic. In a Chinese-language discussion, choose what carries over and take the resulting prompt to your image tool.

To write an image prompt from a reference with AI, start with the reason someone would spend time looking at the original. A rail map might draw a reader in by arranging possibilities along a route. A puzzle asks them to find something. The fafa-hook skill helps you carry that idea into a different subject, with as much or as little visual resemblance as suits your project.

The conversation and the finished model prompt are in Chinese. The writing meant to appear in the image has its own language setting. If you are planning a word puzzle, reserve space for the grid and lay out the letters separately. A misplaced letter can make an otherwise attractive puzzle impossible to solve.

What it does

  • Extend an illustrated series using artwork you have permission to adapt.
  • Turn a reference chart into a layout brief for a different set of figures.
  • Decide whether a new edition keeps its predecessor’s palette or just its organising idea.
  • Specify one language for the visible headings and labels.
  • Reserve a blank area for a supplied logo to be added during layout.

Install and use

  1. Install the skill in an image-capable agent. You can draft the prompt without an image service account.
  2. Supply your reference and topic, then agree in Chinese on the content, visual resemblance and intended format.
  3. Copy the final prompt into your chosen image tool. Add exact lettering or an approved logo during production.

FAQ

Source

Made by 发发 (fafa). Original skill: fafa-hook, published on GitHub under the MIT licence.

View on GitHub