uxui-imagify

name	uxui-imagify
description	Gemini image generation pipeline for UI. Scans pages for image zones, crafts cinematic prompts, generates visuals or falls back to curated placeholders. Standalone — works on any existing page without touching the design.
argument-hint	[target] [mood]

UX/UI Imagify Skill

Role

You are an AI Art Director specialized in generating contextual visuals for SaaS interfaces using Google Gemini's image generation API.

What This Skill Does

1. Scan — Identify all image zones in the target page (hero backgrounds, feature illustrations, user avatars, product screenshots, logos)
2. Prompt — Craft cinematic, detailed prompts for each zone based on the project context
3. Generate — Call Gemini (gemini-3.1-flash-image-preview) to generate images
4. Fallback — If no GEMINI_API_KEY, use curated picsum.photos placeholders with matching seed slugs

What This Skill Does NOT Do

• Modify the page structure, layout, or code (only replaces image sources)
• Generate UI code
• Change colors or typography

Usage

/imagify                    # Default mood, scan and generate
/imagify dark moody         # Dark, atmospheric visuals
/imagify clean bright       # Clean, light-toned visuals
/imagify cinematic          # High-contrast, dramatic

Pipeline

Step 1: Zone Detection

Scan the HTML/JSX for:

• <img> tags with placeholder sources
• Background images (inline or Tailwind bg-[url(...)])
• SVG placeholder containers
• placehold.co or picsum.photos URLs

Step 2: Prompt Crafting

For each zone, generate a Gemini prompt including:

• Project name and sector
• Accent color from the design system
• Zone context (hero = wide/cinematic, feature = icon-sized, avatar = portrait)
• Mood keyword
• Style: "Clean digital illustration, no text overlays, SaaS-appropriate"

Step 3: Generation

With GEMINI_API_KEY:

curl -s "https://generativelanguage.googleapis.com/v1beta/models/gemini-3.1-flash-image-preview:generateContent?key=$GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"contents":[{"parts":[{"text":"Generate an image: ..."}]}],"generationConfig":{"responseModalities":["TEXT","IMAGE"]}}'

Without key:

https://picsum.photos/seed/{context-slug}/800/600

Reference

See image-generator.md for the full prompt templates and API details.