opencode-skill/skills/image-generation/SKILL.md

name, description

name	description
image-generation	Generate images from text prompts using Chutes AI image generation. Use this skill when the user wants to create AI-generated images from descriptions.

Image Generation

Generate AI images from text prompts via python3 scripts/image_gen.py generate <prompt> [options].

Commands

Command	Args	Description
generate	<prompt> [--width W] [--height H] [--steps N] [--seed N]	Generate image from prompt

Options

Option	Default	Range	Description
--width	1024	576-2048	Image width in pixels
--height	1024	576-2048	Image height in pixels
--steps	9	1-100	Number of inference steps
--seed	null	0-4294967295	Random seed (null = random)
--guidance-scale	0	0-5	Guidance scale for generation
--shift	3	1-10	Shift parameter
--max-seq-len	512	256-2048	Max sequence length

Examples

# Basic generation
python3 scripts/image_gen.py generate "a high quality photo of a sunrise over the mountains"

# Custom dimensions
python3 scripts/image_gen.py generate "a futuristic city at night" --width 1280 --height 720

# With seed for reproducibility
python3 scripts/image_gen.py generate "a cute cat sitting on a windowsill" --seed 42

# High quality with more steps
python3 scripts/image_gen.py generate "a detailed portrait of a woman in renaissance style" --steps 20

Workflow

1. Run generate with your prompt
2. Script saves image as generated_[timestamp].png
3. Returns image path: generated_1234567890.png [12345]

Output Format

• Success: Image saved: filename.png [id]
• Error: Error: message (to stderr)
• Images saved to current working directory as PNG files

Notes

• Requires CHUTES_API_TOKEN in environment
• Prompt length: 3-1200 characters
• Large images (2048x2048) take longer to generate
• Images are saved locally, not returned as base64 to save memory