---

name: stitch description: AI-powered UI and image generation using Gemini 3 (Nano Banana Pro)

Stitch Skill

Generate production-ready UIs, components, and images using Google's Gemini 3 AI.

Configuration

Requires GEMINI_API_KEY in environment variables.

Models Used

• gemini-3-pro-preview: Code/UI generation with advanced reasoning
• gemini-3-pro-image-preview: Pro image generation (Nano Banana Pro) - up to 4K
• gemini-2.5-flash-image: Fast image generation (Nano Banana)

Actions

UI/Code Generation

generate_ui

Generate a complete UI from a text description.

• prompt (str): Description of the UI
• format (str): "html", "react", or "css"

generate_component

Generate a reusable UI component.

• prompt (str): Component description
• framework (str): "react", "vue", "svelte", or "html"

Image Generation

generate_image

Generate an image from text.

• prompt (str): Image description
• aspect_ratio (str): "1:1", "16:9", "9:16", etc.
• resolution (str): "1K", "2K", or "4K"
• output_path (str): Path to save the image

edit_image

Edit an existing image.

• prompt (str): Edit instructions
• image_path (str): Path to source image
• output_path (str): Path to save edited image

generate_logo

Generate a professional logo.

• prompt (str): Logo description
• output_path (str): Path to save

generate_infographic

Generate an infographic with Google Search grounding.

• prompt (str): Infographic topic
• output_path (str): Path to save

Examples

Copy# Generate UI code
result = skill.execute("generate_ui", prompt="a modern dashboard with sidebar", format="react")

# Generate an image
result = skill.execute("generate_image", 
    prompt="futuristic city at sunset", 
    aspect_ratio="16:9", 
    resolution="2K",
    output_path="city.png")

# Generate a logo
result = skill.execute("generate_logo",
    prompt="coffee shop called 'The Daily Grind'",
    output_path="logo.png")