Dunhuang Aura Skill
Skill for commercial visuals in the spirit of Dunhuang mineral pigments: 5:2 covers, posters and product hero images.
- 109 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The skill writes prompts and checks them with a script; generation runs in the host service.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/dunhuang-aura-skillDetects 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 writes prompts and checks them with a script; generation runs in the host service.
This is third-party code. Review the repository files before installing.
What it does
Dunhuang Aura Skill creates and edits images in one style: charcoal cave rock, sandy fresco panels, cinnabar, malachite and lapis green and blue, restrained ochre light. It is more than a prompt collection: it determines the deliverable, ratio, text state and composition, then constrains palette, light, object geometry and element density. It supports 5:2 covers for X, full-frame images without text, article covers with an exact title, ad visuals, event posters, phone and product shots, and edits of existing images: remove text, fill the background, balance the sides. The `scripts/check_prompt.py` script checks prompt completeness.
Who it is for. Designers and brand owners who need covers and hero images in a recognizable Dunhuang palette.
Good fit when
- An article about AI tools needs a 5:2 cover without text.
- A tea brand hero image is made while keeping the jar shape.
- All inscriptions are removed from a finished frame and the gap is filled.
- A prompt is checked with the script before it goes to the generator.
Not a fit when
- A reconstruction of paintings or a religious subject is needed: era, cave and sources must come first.
- The style does not suit the brand.
- The host has no image generation.
Example request
Use $dunhuang-aura to make a text-free 5:2 cover about AI image generation, with a large subject at the edges and a water strip in the middle.Limitations
The description and instructions are in Chinese. The skill targets commercial visuals and by default excludes Buddha images, bodhisattvas, flying apsaras, worship scenes and copies of fresco figures. The image itself is made by host tools. `check_prompt.py` only checks prompt completeness and does not replace viewing the result. Last commit in September 2026. The result depends on the host image model, and exact rendering of logos and lettering remains a generator weak spot, so final text is often added separately.
How to disable. Delete the `dunhuang-aura` directory from the client skills folder (`~/.codex/skills` or `.claude/skills`).
FAQ
Which formats are supported?
5:2 covers for X, full-frame images without text, article covers with a title, ad visuals, posters and device shots.
What does the check script do?
`check_prompt.py` checks ratio, text ban, palette, composition, materials, light and template words. It does not judge the generated result.
Are religious images included?
No. Buddha, bodhisattvas, flying apsaras and worship scenes are excluded by default, and frescoes serve as a source of colour and form.
How do I run the tests?
Run `python3 -m unittest discover -s tests -v` from the repository root.
How do I edit a finished image?
Ask to remove the text, keep the product, light, palette and composition, and fill the empty area with a cave gallery. The skill handles such edits as a separate workflow.
Discussion
No questions or reviews yet.
Related
SkillsNuyoah Image Reverse Prompt
nuyoah-image-reverse-prompt
A skill that breaks a picture into frame, light, color and materials and assembles a ready Chinese prompt for image generation.
Five image-generation skills: Qwen-Image prompts, logos with shared geometry, diary comics, transparent PNGs and process dioramas
A skill that turns one sentence or a reference image into a 3D holographic card web page and an editable Blender project
Skill for animation in painting styles: 36 art styles and 9 explainer grammars, rendered by code with uv, Playwright and ffmpeg.
Qwen-Image 2.1 Prompter
qwen-image-2.1-skill
A skill that rewrites prompts for Qwen-Image-2.1 by Alibaba's official rules: text-to-image and multi-image editing
A skill that turns a picture into a generator prompt: frame breakdown, an enhanced version, three variants and platform editions.
README in short
The style is described as a system: palette, materials, light and three depth layers, and a script catches vague prompts.