Agent skills for generating videos with Seedance 2.0 and Seedance 2.0 fast models.
Note: These skills are built for
https://seedance.infinitext.cn/. They also work with the official Volcengine Ark API, but compatibility is not guaranteed.
npx skills add mercallureAI/seedanceThen select the skill(s) you want and the agent(s) to install them on.
For use with the meme-master CLI. Covers all CLI commands (video generate, video list, video get), every generation mode, flags, validation rules, and prompt techniques.
npx skills add mercallureAI/seedance --skill seedanceFor direct API usage (curl, Python SDK, etc.) without any CLI wrapper. Covers all REST endpoints, request/response schemas, async polling, and code recipes.
npx skills add mercallureAI/seedance --skill seedance-apiBoth skills are fully self-contained and can be installed independently.
skills/
├── seedance/ CLI-based (meme-master)
│ ├── SKILL.md Core skill instructions
│ └── references/
│ ├── cli-reference.md Complete flag & command reference
│ ├── input-specs.md Input constraints & validation rules
│ ├── examples.md CLI recipes for every generation mode
│ └── prompt-tips.md Advanced prompt techniques
│
└── seedance-api/ Direct API (curl / Python SDK)
├── SKILL.md Core skill instructions
└── references/
├── api-reference.md Endpoint specs & response schemas
├── input-specs.md Input constraints & resolution tables
├── examples.md Code recipes for every generation mode
└── prompt-tips.md Advanced prompt techniques
These skills work with any agent that supports the Agent Skills format, including Claude Code, Cursor, Codex, Gemini CLI, and others.
You need an ARK_API_KEY from the Volcengine Ark console. Set it as an environment variable:
export ARK_API_KEY="your-key-here"No credentials are stored in or required by these skill files.
| Model | Model ID | Best For |
|---|---|---|
| Seedance 2.0 fast | doubao-seedance-2-0-fast-260128 |
Speed & cost efficiency |
| Seedance 2.0 | doubao-seedance-2-0-260128 |
Maximum quality |
Both models support: text-to-video, image-to-video (first frame, first+last frame), multimodal reference, video editing, video extension, audio-synced generation, and web search enhancement.
MIT