Loading...
Loading...
Generate images using Google Gemini AI with text prompts and reference images. Use when creating game assets, concept art, UI mockups, promotional images, or any visual content. Supports text-to-image, image-to-image with style transfer, and multiple output sizes. Requires GEMINI_API_KEY environment variable. Triggers on requests for AI image generation, concept art, visual assets, or Gemini images.
npx skill4agent add ccalebcarter/purria-skills gemini-image-generator# Navigate to scripts directory
cd scripts
# Create virtual environment
python3 -m venv venv
# Install dependencies
./venv/bin/pip install -r requirements.txt # Unix
# or
.\venv\Scripts\pip install -r requirements.txt # Windows
# Set API key
export GEMINI_API_KEY="your-api-key" # Unix
# or
$env:GEMINI_API_KEY = "your-api-key" # PowerShellpython generate.py --prompt "A serene mountain landscape at sunset" --output landscape.pngpython generate.py --prompt "Same scene but in winter" --reference landscape.png --output winter.png[Subject] + [Style] + [Composition] + [Technical] + [Mood]"A bio-mimetic robot with Art Nouveau brass gears and botanical vine patterns,
centered composition on transparent background, flat vector style suitable for
game sprite, warm golden hour lighting, whimsical and charming mood"| Parameter | Required | Description |
|---|---|---|
| Yes | Text description of desired image |
| Yes | Output file path (.png) |
| No | Reference image for style guidance |
| Error | Solution |
|---|---|
| API key not valid | Check GEMINI_API_KEY is set correctly |
| 403 Forbidden | API key may have IP restrictions |
| Model not found | Model names change; check Google AI docs |
| No image generated | Try simpler prompt, check API quota |
game-assets-team