Use the reference image only as high-level inspiration for composition and overa...

Use the reference image only as high-level inspiration for composition and overall quality. Do not copy the layout exactly. Design a next-generation professional creative application for AI image generation called a Visual Prompt Compiler. The interface should resemble premium creative software such as Figma, Unreal Engine, Adobe Substance, or DaVinci Resolve rather than a traditional dashboard. The design uses a sophisticated dark mode with deep obsidian backgrounds, slate-gray panels, subtle glass accents, soft shadows, generous spacing, minimalist sans-serif typography, and a high signal-to-noise ratio. Every component should have a clear purpose with exceptional visual hierarchy. The application is organized into three primary workspaces. Left Panel — Creative Preset Library A polished asset browser containing curated one-click creative presets displayed in a clean two-column grid. Instead of representing prompts, presets represent creative intent. Examples include: Portrait Studio Character Design Product Photography Editorial Fashion Cinematic Poster Environment Concept Fantasy Character T-Shirt Design YouTube Thumbnail Each card contains a high-quality thumbnail, subtle hover animations, category iconography, and lightweight metadata. The library feels curated rather than cluttered. Center Panel — Visual Reasoning Pipeline The center workspace is the heart of the application. Instead of feeling like a form, it feels like an intelligent creative reasoning engine. Rename the section to something such as: Visual Reasoning Scene Assembly Semantic Composer Creative Reasoning The workspace is organized as a vertical six-stage pipeline grouped into logical sections. Creative Direction Scene Subject Cinematography Camera Lighting Style Rendering Final Look Each stage is visually unique instead of using identical cards. Examples: Scene uses a landscape/environment thumbnail. Subject shows a character bust. Camera displays a lens visualization. Lighting contains a lighting diagram. Rendering displays material/style samples. Final Look shows a small before/after mood comparison. Each stage contains: selected semantic chips dropdown selectors visual pills expandable advanced controls No text prompt editing is visible. The currently selected stage glows with a premium animated accent. Relationship Intelligence Every stage includes a dedicated Relationship Intelligence section. Rather than generic recommendations, it explains why decisions are being suggested. Examples: "35mm is recommended because your subject distance and framing produce a more natural perspective." "Back lighting complements the selected sunset environment." "A standing pose improves silhouette readability at the current camera distance." Changing one stage subtly updates affected stages. For example: Changing Camera briefly highlights: Subject Lighting Final Look Changing Subject updates: Camera Composition Pose Small animated dependency indicators briefly pulse to show that related semantic relationships have been recalculated. The interface should feel alive and intelligent rather than static. At the bottom of the workspace is a large Creative Freedom slider balancing Precision and Creativity. Right Panel — Semantic Subject Editor The right panel is no longer a generic property inspector. It is a dedicated Semantic Subject Editor. At the top is a segmented toggle: Subject Properties Compiled Prompt Subject Properties is the default. Interactive 3D Preview A clean wireframe human mannequin appears inside a viewport. The mannequin is neutral, unclothed, and displays only proportions and pose. No hairstyle, clothing, or accessories. It exists only for: body proportions pose camera composition spatial placement The viewport includes subtle transform controls and rotation controls. The mannequin should feel interactive rather than decorative. Semantic Attribute Groups Instead of one long property list, organize controls into collapsible sections. Identity Gender Age Height Build Body Type Pose & Expression Pose Head Rotation Eye Direction Facial Expression Hand Position Weight Distribution Composition Camera Facing Direction Character Scale Distance From Camera Silhouette Appearance Physical Features Controls use elegant segmented buttons, sliders, dropdowns, and chips. Changing any property causes the corresponding Subject stage in the center reasoning pipeline to softly pulse, reinforcing the connection between both workspaces. Compiled Prompt View Switching to Compiled Prompt replaces the inspector with the fully assembled prompt generated from the semantic graph. The prompt appears as a beautifully formatted structured document with syntax highlighting, semantic grouping, and copy/export controls rather than a plain text box. Generate Section Anchor a large premium Generate Image button across the entire bottom of the right panel using a vibrant blue-to-violet gradient. Above the button display a compact semantic readiness indicator such as: Semantic Quality: 97% No relationship conflicts detected 2 AI recommendations available This reinforces confidence before generation. Visual Language Use distinct semantic accent colors to categorize stages. For example: Creative Direction → Cyan Cinematography → Blue Style → Violet Active elements glow softly with refined neon lighting while maintaining a professional aesthetic. Avoid visual clutter. The interface should feel spacious, intelligent, elegant, and highly interactive. The final result should look like a premium creative operating system for AI image generation—not a prompt editor, but a visual semantic reasoning environment that intelligently assembles scenes through interconnected creative decisions.