A Claude skill that turns you into a Seedance 2.0 power user — with expert guidance on multimodal prompts, camera replication, video extension, audio design, beat-syncing, and more.
Download the seedance-2.0.skill file to your computer.
Go to claude.ai and click your profile icon → Settings.
In the left sidebar, click Skills (or Custom Skills, depending on your plan).
Click Install Skill (or drag and drop the .skill file into the upload area).
The skill will appear in your skills list as seedance-prompting. It's now active.
✅ No restart needed — the skill is available immediately in any new conversation.
Once installed, Claude will automatically apply this skill whenever you ask about Seedance 2.0 video generation. You don't need to invoke it manually.
- "Write me a Seedance 2.0 prompt to extend this video by 10 seconds"
- "How do I clone the camera movement from a reference video in Seedance?"
- "Create a one-shot action sequence prompt for Seedance with 3 location images"
- "Help me write a beat-synced prompt using these 6 photos"
- "I want to replace a character in my video using Seedance"
Claude will respond with structured, ready-to-paste prompts and guidance tailored to your request.
| Capability | What You Can Ask |
|---|---|
| Multimodal inputs | How to combine images, videos, audio, and text |
| Camera replication | Clone camera movements from a reference video |
| Character consistency | Keep faces, outfits, and props consistent across cuts |
| Video extension | Seamlessly add seconds to an existing video |
| Video editing | Swap characters, change hair, alter plot |
| Creative effects | Replicate transitions, particles, and visual FX |
| Audio design | Reference voice tone, music, ambient sound |
| One-shot filming | Continuous camera with no cuts |
| Beat-synced edits | Sync cuts and transitions to music rhythm |
| Story completion | Generate narrative from storyboard or mood images |
When writing Seedance prompts, always use @asset-name to reference your uploaded files:
@image1 as the first frame
Reference @video1 for camera movement
Use @audio1 as background music
Extend @video1 by 6s
For video extension: set the generation length to the new portion only, not the total duration.
Entry points:
- First/Last Frame — single image + text prompt
- Omni-Reference — any mix of image + video + audio + text (use this for multimodal)
| Type | Max Files | Max Duration / Size |
|---|---|---|
| Images | 9 | 30 MB each |
| Videos | 3 | 2–15s total · 50 MB each |
| Audio | 3 | 15s total · 15 MB each |
| Total files | 12 | across all types combined |
- Be explicit — always tell the model what each
@assetis for - Use time segments for complex scenes:
0–3s: ... 4–8s: ... - State "no cuts" clearly if you want a one-shot result
- Realistic human faces cannot be uploaded — use illustrated or stylized characters
- Seedance 2.0 Fast is available if servers are slow — same quality, shorter wait
Go to Settings → Skills, find seedance-prompting, and click Remove.