| name | generating-images |
| description | MUST be used when you need to generate images from text prompts using AI. Creates high-quality images with Gemini Imagen using customizable prompts, aspect ratios, and sizes. Perfect for creating artwork, illustrations, concept art, posters, banners, profile pictures, or any visual content. Triggers on: generate image, create image, make an image, image from prompt, AI art, generate picture, create artwork, make illustration, imagen, text to image, draw an image, visualize this, create visual, generate graphic, make poster, create banner. |
Generating Images
Generate high-quality images from text prompts using Gemini's Imagen model.
When to Use
- User wants to create an image from a description
- Need artwork, illustrations, or concept art
- Creating posters, banners, or marketing materials
- Generating profile pictures or avatars
- Visualizing ideas or concepts
- Creating custom graphics for any purpose
Available Commands
This skill is invoked through the anima skill runner — anima skill run handles
the working directory and env injection so there's no cd dance.
generate — Generate a PNG via Gemini 3 Pro Image (synchronous; typically 10–30s)
Inspect:
anima skill help generating-images generate
Usage
anima skill run generating-images generate "<prompt>" <output-path> [options]
Arguments
<prompt> — Text description of the image to generate (required, quoted)
<output-path> — Absolute path where the PNG will be written (required)
--aspect-ratio <ratio> — Image aspect ratio (optional, default: "1:1")
- Options: "1:1", "16:9", "9:16", "4:3", "3:4", "2:3", "3:2"
--size <size> — Image resolution (optional, default: "2K")
- Options: "1K", "2K", "4K"
Examples
anima skill run generating-images generate "A serene mountain lake at sunset" ~/images/lake.png
anima skill run generating-images generate "Futuristic cityscape at night" ~/images/city.png --aspect-ratio 16:9 --size 4K
anima skill run generating-images generate "Portrait of a wise elder" ~/images/portrait.png --aspect-ratio 2:3
anima skill run generating-images generate "Abstract geometric patterns" ~/images/wallpaper.png --aspect-ratio 9:16 --size 4K
How It Works
- Takes your text prompt
- Sends it to Gemini 3 Pro Image model
- Generates high-quality image based on the description
- Saves as PNG to your specified location
- Reports file size and success
Requirements
- GEMINI_API_KEY environment variable must be set
- Bun runtime
Tips for Good Prompts
- Be specific and descriptive
- Include style preferences (photorealistic, illustrated, watercolor, etc.)
- Mention lighting, mood, and atmosphere
- Describe composition and perspective
- Add details about colors and textures
Example Prompts
Photography style:
"Professional photograph of a cozy coffee shop interior, warm lighting, wooden tables, plants in corners, soft bokeh background, shot with 35mm lens"
Illustration style:
"Whimsical watercolor illustration of a garden with butterflies, soft pastel colors, dreamy atmosphere, children's book style"
Abstract/Artistic:
"Abstract geometric composition with flowing curves, gradient from deep purple to electric blue, modern minimalist design"
Concept Art:
"Sci-fi space station interior, holographic displays, metallic surfaces, dramatic lighting through large windows showing stars, cinematic composition"
Notes
- Images are generated at high quality (2K or 4K resolution)
- Aspect ratios match common use cases (social media, prints, screens)
- PNG format preserves quality and supports transparency
- Generation typically takes 10-30 seconds depending on complexity
- Each generation uses Gemini API credits
- The script auto-creates output directories if they don't exist