AI Rendering Prompt Generator

Problem

Creating render prompts often hits a language gap: a user knows they want 'a light-luxury bedroom with warm lighting and wood textures,' but the prompt sent to Midjourney, Stable Diffusion, or DALL-E may lack style, materials, lighting, and platform parameters, and may also miss negative prompts. This skill structures the conversion from Chinese descriptions into professional English prompts, reducing manual keyword translation and platform parameter lookup.

How It Works

It works around three core data assets:
- data/scenes.json: 20 scene presets covering residential, commercial, outdoor, and architectural exterior use cases
- data/keywords.json: an 89-item Chinese keyword library covering materials, furniture, lighting, and more
- templates/prompt_templates.json: prompt templates for organizing style, atmosphere, and output format

Given a short Chinese description, the skill matches a scene and style, such as 'modern minimalist living room' in residential spaces or 'trending café' in commercial spaces, then converts Chinese keywords into English expressions. Built-in styles include Modern Minimalist, Nordic, New Chinese, Light Luxury, Industrial, Japanese, American, Tech, French, and Eclectic. Atmosphere presets include warm comfort, luxury, fresh natural, futuristic, and other moods. Multi-platform output adapts to common parameter formats and can suggest aspect ratios, quality levels, and negative prompts, such as 16:9 for full interior views, 3:2 for building exteriors, lower quality for quick sketches, and higher quality settings for final renders.

Boundaries and Caveats

This skill is useful for quickly producing paste-ready render prompts, especially for Chinese-speaking users, interior concepts, commercial spaces, outdoor landscapes, and building exteriors. If the project depends heavily on brand assets, real floor plans, precise dimensions, specific character designs, or complex spatial constraints, additional references, drawing information, and clearer structural descriptions are still needed. Style and atmosphere are directional controls, not guarantees that a model will reproduce a specific project exactly. Even after generation, it is still advisable to use negative prompts to suppress low-quality artifacts and compare multiple prompt variants.