| name | seedream-5-image |
| description | Use when generating images with Seedream 5.0, including text-to-image, image-to-image, multi-image inputs, or local image files that need automatic URL upload first. |
Seedream 5.0 Image Skill
This skill can:
- generate images from text prompts
- accept one or more image inputs
- save numbered output files when Seedream returns multiple generated images
- auto-upload local image files through the configured Qiniu bridge
- call Seedream 5.0 directly with Ark auth
Usage
python3 scripts/generate_seedream.py "Jazz Festival poster"
python3 scripts/generate_seedream.py "poster redesign" --image https://example.com/a.png
python3 scripts/generate_seedream.py "poster redesign" --image /tmp/local.png
python3 scripts/generate_seedream.py "multi reference" --image https://a.png --image /tmp/b.png --sequential --max-images 3
Requirements
config.json in skill root or ~/.seedream-config.json (contains ARK_API_KEY and Qiniu settings)