- name
- jpeng-image-generator
- description
- AI image generation skill using DALL-E, Stable Diffusion, or Midjourney API. Generate, edit, and vary images from text prompts.
- version
- 1.0.0
- author
- jpeng
- tags
- ["image", "ai", "dalle", "stable-diffusion", "generation"]
Image Generator
Generate AI images using DALL-E, Stable Diffusion, or other providers.
When to Use
- User wants to generate an image from text
- Create variations of existing images
- Edit images with AI inpainting
- Generate thumbnails or illustrations
Configuration
# OpenAI DALL-E
export OPENAI_API_KEY="sk-xxx"
# Stability AI
export STABILITY_API_KEY="sk-xxx"
# Replicate (for SDXL, etc.)
export REPLICATE_API_TOKEN="r8_xxx"Usage
Generate image
python3 scripts/generate_image.py \
--prompt "A serene mountain landscape at sunset" \
--size "1024x1024" \
--output "./output.png"Generate variations
python3 scripts/generate_image.py \
--vary "./input.png" \
--count 4 \
--output-dir "./variations"Edit image (inpainting)
python3 scripts/generate_image.py \
--edit "./input.png" \
--mask "./mask.png" \
--prompt "Add a rainbow in the sky"With style presets
python3 scripts/generate_image.py \
--prompt "A futuristic city" \
--style "cyberpunk" \
--negative "blurry, low quality"Output
{
"success": true,
"image_path": "./output.png",
"revised_prompt": "A serene mountain landscape..."
}