surreal-pop-collage skill
An image skill that turns a photo into a surreal pop collage: a black-and-white anchor and one impossible giant object.
- 360 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The skill only composes a text prompt; the photo then goes to the image model you choose
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/surreal-pop-collageDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill only composes a text prompt; the photo then goes to the image model you choose
This is third-party code. Review the repository files before installing.
What it does
surreal-pop-collage, a skill with its protocol in `SKILL.md`, assembles an image-model prompt on a four-part scheme. The source photo is desaturated and stays a recognizable anchor, the background is replaced by two or three huge flat colour shapes with colours drawn from the picture itself. Then one impossible giant element is chosen, growing out of the scene. The rules in `SKILL.md` are strict: only one object, no gradients or volumetric shadows, no text or watermarks, no stock red suns or dolphins. For cultural landmarks the giant comes from the place's legend. `examples/` holds samples for a landscape, a city and a portrait.
Who it is for. Designers and social media authors who want a recognizable cover and illustration style built from their own photos.
Good fit when
- You need to turn a city photo into a pop-collage poster.
- A series of covers in one visual language is being prepared.
- The model result came out overloaded and a rule check is needed to correct it.
- You work in a generator with no agent and need to write the prompt by hand from the scheme.
Not a fit when
- You need realistic retouching or photorealism.
- You want several impossible elements in the picture.
- You need a finished file rather than a prompt: the skill does not draw the picture itself.
Example request
Turn this photo into a surreal pop collage: define the anchor, colour shapes and one giant object, then produce the prompt.Limitations
The skill produces a prompt, and result quality depends on the image model chosen. The documentation is in Chinese. An OpenAI agent description sits alongside in `agents/openai.yaml`. The last commit is dated August 2026. The style is strict and narrow, so every picture will resemble its neighbours. The rules do not guarantee that an image model follows them on the first try. If something misbehaves, check the README and the repository's open discussions.
How to disable. Delete the skill folder from the agent skills directory, for example `~/.kimi-code/skills/surreal-pop-collage/`.
FAQ
Is installation needed?
For agents, copy `SKILL.md` and `agents/` into the skills directory. For ChatGPT, Midjourney or Jimeng you can just write a prompt from the README recipe.
Why only one giant object?
The "one thing is wrong" rule: a second impossible element should be removed or the surreal effect is lost.
Where do the colours come from?
From the photo itself: colours are purified, complemented or inverted by mood.
What is the prompt format?
Four parts: collage and format, a desaturated anchor, flat colour shapes, and an impossible element with small details along an arc.
Does it suit portraits?
`examples/` has three types: landscape, city and person.
Discussion
No questions or reviews yet.
Related
SkillsA skill that turns a picture into a generator prompt: frame breakdown, an enhanced version, three variants and platform editions.
Five image-generation skills: Qwen-Image prompts, logos with shared geometry, diary comics, transparent PNGs and process dioramas
A Codex skill that turns a photograph into a vertical composition with an abstract memory panel and an English title
Codex skill that turns a portrait, photo or text description into a retro illustration in the style of an early CRT interface.
Two image skills for wordless visuals: sparse abstract geometry on matte paper and white chalk on a green board, one idea each.
A skill that makes an editorial poster from a photo: the shot stays unchanged, with an abstract panel and a local title beside it.
README in short
The style rules are boiled down to five tenets: monochrome reality, flat dreams, one wrong thing, elements from the shot, colours from the shot.