
Seedance Camera Direction
Writes precise, physically plausible camera moves for Seedance 2.0 video prompts
- Repository
- Emily2040/seedance-2.0
- Stars
- 8,008
- Licence
- MIT
- Files
- 1 file
What it does
This skill teaches an AI agent to describe camera work for AI video generation with the Seedance 2.0 model. When you ask for a dolly, pan, handheld shot, crane move or similar, the agent picks one clear camera move, states its start, speed, subject relationship and endpoint, and writes it as a ready-to-use prompt sentence. It also avoids stacking conflicting moves and keeps camera direction consistent across continuation clips and reference videos.
Great for
- Creators writing text or image-to-video prompts for Seedance 2.0 who want cinematic camera language
- Filmmakers or hobbyists planning shot lists who need professional shot terms translated into prompt wording
- Developers building prompt tools on SKILL.md-compatible agents who want consistent camera phrasing
Skip it if
- People who only need general video editing or real camera operation, since this only shapes generation prompts and does not edit footage
- Users who expect the output to guarantee a specific result, because the video model may not follow the wording exactly
Try asking
- “Give me a slow push-in camera move for a character reading a letter, ending on a tight close-up”
- “Write a locked macro shot prompt for a watch mechanism with the second hand ticking”
- “Turn this storyboard note into a single camera phrase for a crane reveal of a city skyline”
- “Keep the camera movement from @Video1 but do not copy the performer or room”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
seedance-camera.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip seedance-camera.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip seedance-camera.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Name one primary camera move per clip and list alternatives separately, since stacked moves like drone plus dolly plus handheld tend to conflict.
- Describe the start frame, speed and endpoint explicitly, and tie the move to a visible subject action.
- For complex camera motion, use a reference video with an instruction that it transfers camera rhythm only.
What's inside
- SKILL.md
Redistributed unmodified from Emily2040/seedance-2.0 under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.