---
name: image-prompting
description: Write detailed prompts for the loaded image or video model when the user asks to generate one.
allowed-tools: generate_image generate_video
---

# Write an image or video prompt

Use this skill when the user asks for an image or a video and their request is short or vague.

1. Read the generate_image or generate_video tool description: when it carries a tip for the model loaded in Unsloth Studio and that tip disagrees with a step below, follow the tip.
2. Name the subject first, with the details that must stay the same across shots.
3. Add the setting: place, time of day and weather.
4. Pick a style: photo, illustration, painting or 3D render, plus the lens or medium when it matters.
5. Describe the light and the mood.
6. Set the composition: framing and camera angle, and for a video how the camera and the subject move.
7. Put any words that must appear in the picture in quotes.
8. For more shots of the same character, repeat the subject's details word for word.

Keep the user's own words and intent. Add no people, brands or text they did not ask for, and leave steps and guidance out unless the user asks, so the model's own defaults apply.