| name | image-generation |
| description | Use this skill when asked to "generate an image", "create a picture", "make an illustration", "draw something", or when visual content needs to be created. Provides image generation via the gen-image CLI tool. |
Image Generation Skill
Use gen-image CLI to generate images with OpenAI's gpt-image-1 model.
Quick Start
gen-image "A cat wearing sunglasses"
gen-image -q low -s square -o /tmp "Your prompt"
Common Options
| Flag | Description | Values |
|---|
-q | Quality | low ($0.01), medium ($0.04), high ($0.17) |
-s | Size | square, portrait, landscape |
-o | Output dir | Path (default: current dir) |
-t | Transparent | Enable transparent background |
-n | Count | Number of images to generate |
Workflow
-
Generate image:
gen-image -q low -s square -o /tmp "Your prompt"
-
Find the output file:
ls -t /tmp/generated_output_*.png | head -1
-
View image (use Read tool on the file path)
-
Send to Discord (use discord_send_file tool)
Tips
- Start with
low quality for testing - upgrade only when needed
- Be specific in prompts - detailed prompts give better results
- Use
/tmp for output - keeps generated images temporary
- Prefer
gpt-image-1 model (DALL-E being deprecated May 2026)
Size Reference
| Alias | Dimensions | Best For |
|---|
square | 1024x1024 | Icons, avatars |
portrait | 1024x1536 | Mobile, posters |
landscape | 1536x1024 | Desktop, banners |
