Colossal Celestial Palace
Turn Chinese scenes into grand celestial art prompts
Colossal Celestial Palace
Turn Chinese scenes into grand celestial art prompts
Showcase
Description
Turn a Chinese scene concept into a Chinese colossal celestial-palace prompt ready for AI image generation. Preserve the number of people, their actions, key objects, location, era constraints, and mood so the image stays true to the original intent while gaining a grand Eastern sense of scale. Whether the scene features tea, music, meetings, banquets, or reading in a palace—or a celestial landscape of cloud seas, colossal pillars, moon gates, and supersized architecture—you get clear relationships between activities, coherent architectural logic, and layered depth between mountains, water, and empty space. The Skill defaults to a Chinese celestial-palace aesthetic inspired by white-jade-and-gold palaces and solitary cities above a sea of clouds. It combines traditional palace order, cinematic widescreen composition, low-angle golden light, and colors such as ivory white, cinnabar, antique gold, and blue-green. It adjusts character scale, viewpoint, lighting, and negative space to fit each scene. It will not let colossal scenery obscure the narrative actions you specify, and it avoids modern cities, neon sci-fi, cluttered text, and common problems with building or character generation. Each run provides a visual design brief, prompts tailored separately for ImageGen and Midjourney 8.2, and a few possible directions for adjustment. It is suited to concept development, storyboard planning, Chinese-style scene creation, celestial-palace architecture visualization, and optimizing Chinese image-generation prompts.
Recommended by
Dancang
Why we love this skill
Precisely turns Chinese scene ideas into image-ready celestial-palace prompts while preserving actions, objects, architectural scale, and spatial order. Optimized separately for ImageGen and Midjourney.
Best for
For creators developing Chinese-style celestial-palace scenes, concept art, or optimized Chinese image-generation prompts.
How to use this Skill
Provide your input
Enter a Chinese scene concept. It can include characters, actions, objects, locations, era, mood, or aspect-ratio requirements.
Run the Skill
The Skill turns your concept into a Chinese colossal celestial-palace design while preserving key narrative relationships and visual requirements.
Review your result
Get a visual design brief plus image-generation prompts for ImageGen and Midjourney, along with a few possible directions for adjustment.
Related Skills
View allClassic of Mountains and Seas
Transform a single scene concept into Classic of Mountains and Seas style mythological image prompts featuring moonlit sea of clouds, celestial palaces, and colossal beasts. Ideal for scenes like beasts descending, ancient people looking up, crowds watching ceremonies, solitary beast portraits, and scenic views. Beast anatomy, colors, and patterns are clearly described based on historical records or user settings to avoid western dragons or generic fantasy creatures. The visuals maintain the oriental textures, fine brushwork, cold-warm lighting, and epic feel of the reference images. Users can specify locations, times, weather, character counts, aspect ratios, and beast actions, or start with a vague idea to flesh out visual details. Outputs can include complete Midjourney English prompts, Chinese explanations, and negative constraints for beast form, character scale, moon disc, color temperature, and architectural style. Sets of prompts with consistent settings can also be generated when needed. For ImageGen, native Chinese prompts without Midjourney parameters are provided. After receiving images, you can refine beast count and structure, character scale, moon position, color, and oriental brushwork based on actual deviations to make subsequent prompts match your expectations.
Desert Megastructure Aesthetic
Turn people, objects, cities, interiors, or other suitable subjects into ready-to-use minimalist still-image prompts in a desert megastructure style. The skill preserves the user’s specified subject, number of people, actions, contact relationships, and scene function, while using low-saturation sand and mineral gray, monolithic wedge-shaped geometric megastructures, hard sunlight, vast pure-dark surfaces, and extreme contrasts of scale to create a silent, restrained, architectural image. It does not rely on a film title as a substitute for specific descriptions of lighting, materials, and composition. You’ll get a clear compositional intent, native Chinese ImageGen prompts, and English Midjourney prompts. You can also reduce the output to a single-platform version as needed. This is suitable for single stills, consistently styled image sets, and correcting common deviations such as oversaturated colors, overly dense machinery, unbalanced scale, or visual noise. The visual language can also be applied to city and interior subjects without forcing in a desert. When you request actual image generation, the skill can additionally check character visibility, contact relationships, style anchors, and consistency across images.

Novel Scene Image Prompts
Extract actual scene cuts one by one from novels, scripts, text storyboards, or plot outlines, transforming them into Chinese empty-scene reference prompts ready for AI image generation. Whether dealing with changes in location, time, weather, or the reappearance of the same location in different plot segments, scenes are independently organized in story order to prevent omissions or mismergers, establishing a stable spatial reference for subsequent comic storyboards. Each scene undergoes specific art design tailored to its narrative function, detailing spatial structure, entrances and exits, movement paths, foreground-midground-background layers, key props, visual focus, lighting relationships, materials, colors, and narrative traces, while classifying them as indoor, outdoor, vehicle, or special spaces. Even with limited source text, necessary visual scene information is supplemented. The final output includes a scene overview, overall visual tone, individual scene designs, Chinese image generation prompts, and Chinese negative prompts. Prompts emphasize cinematic high-angle oblique shots, high-angle wide-angle panoramas, 18mm-24mm lenses, deep depth of field, and complete spatial layouts, strictly maintaining empty scenes without characters or animals, making them ideal for cinematic establishing reference images rather than character storyboards, architectural floor plans, or ordinary interior renderings.
Information
- Version
- v2
- Last updated
- Runtime credits
- Usage-based
- Models
- Auto
- Use cases
- Writing & contentImages & design
- What you get
- BriefPrompt