image-generation
Installation
SKILL.md
Image Generation & Editing
This skill supports two providers — pick whichever fits the task at hand. Both can do text-to-image generation and image editing (including multi-reference composition).
| Provider | Models | Best for |
|---|---|---|
| Google Gemini (Nano Banana) | gemini-3.1-flash-image-preview, gemini-3-pro-image-preview |
Aspect-ratio based output, text-in-image (especially non-Latin scripts), composition with up to 11–14 reference images. |
| OpenAI (gpt-image-2) | gpt-image-2 |
Pixel-precise sizing, mask-based localized editing, generating multiple variants in one call (--n), output formats beyond PNG (jpeg/webp). |
If the user doesn't specify, default to Gemini Nano Banana (gemini-3.1-flash-image-preview) — fastest and cheapest for everyday tasks.
Prerequisites
- Python 3.10+ must be installed.
- The relevant API key is assumed to already be available from the OS environment or a nearby
.envfile:GEMINI_API_KEYfor Google Gemini scriptsOPENAI_API_KEYfor OpenAI scripts