AI image generation with style presets and multi-image support. Powered by Google Gemini.
/plugin install https://github.com/jcasimir/compound-engineering-plugin?plugin=imagegenSet your Gemini API key:
export GEMINI_API_KEY="your-api-key-here"Create the Python environment:
python3 -m venv ~/.gemini-venv
~/.gemini-venv/bin/pip install google-genai Pillow/imagegen "A mountain sunset" sunset.jpg- Text-to-image: Generate images from natural language prompts
- Image editing: Modify existing images with natural language
- Style presets: 23 built-in styles (anime, photorealistic, cinematic, etc.)
- Multiple variations: Generate up to 4 images at once
- Flexible sizing: Multiple aspect ratios and resolutions (1K, 2K, 4K)
| Category | Styles |
|---|---|
| Photography | photorealistic, portrait, product, street, cinematic, aerial |
| Art | anime, watercolor, oil-painting, sketch, comic, pixel-art, 3d-render |
| Era/Aesthetic | 80s-retro, 90s-vintage, minimalist, maximalist, vaporwave |
| Technical | blueprint, infographic, ui-mockup |
| Mood | dark-moody, bright-airy, golden-hour |
# Basic generation
/imagegen "A coffee shop" shop.jpg
# With style preset
/imagegen "A city street" street.jpg --style cinematic
# Multiple variations
/imagegen "A logo" logo.jpg --count 4 --style minimalist
# High resolution
/imagegen "Epic landscape" landscape.jpg --aspect 16:9 --size 2K| Script | Purpose |
|---|---|
generate_image.py |
Text-to-image generation |
edit_image.py |
Edit existing images with prompts |
compose_images.py |
Combine multiple reference images |
multi_turn_chat.py |
Iterative refinement via chat |
- App icons and illustrations
- Product photography mockups
- UI design assets
- Marketing visuals
- Concept art and storyboards
- Logo exploration