Seedance 2.0 Prompt Writing Skills
A skill that teaches an agent to write prompts for ByteDance's Jimeng Seedance 2.0 video model: @ references, camera language, timing, templates
- 4,376 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The skill is made of text instructions: it runs nothing and makes no network calls
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/seedance-2-0-prompt-writing-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill is made of text instructions: it runs nothing and makes no network calls
This is third-party code. Review the repository files before installing.
What it does
The skill turns an agent into a prompt writer for Jimeng Seedance 2.0, ByteDance's multimodal video model. The agent first clarifies the goal of the clip and the list of assets, then assigns each file a role and builds the prompt by a formula: subject, scene, action, camera, timing, transitions, sound and style. The core is the @ reference system: @Image1, @Video1, @Audio1 and so on. The guide shows how to set a first frame, a character's look, a camera move, rhythm or music. For clips longer than 10 seconds it advises splitting the prompt into segments such as 0–3 s and 3–6 s, and camera language gets its own reference: push in, orbit, tracking, handheld, shot sizes from wide to extreme close-up. Ten common tasks are covered, among them character consistency, replicating camera moves, extending and editing a finished video, beat matching, dialogue, one-take shots and educational scenes. The end collects common mistakes and four templates: a 15-second ad, a short drama, a dance video and a scenery montage with music.
Who it is for. For creators of ads, music videos and story clips who make video in Seedance 2.0 and want predictable output from the first tries.
Good fit when
- You have 3–4 reference images and a clip with camera movement, and need each file's role laid out in the prompt
- You need a 15-second product ad built from a photo with segment-by-segment timing
- A camera move or editing style from someone else's video must be replicated on your own material
- A finished video should be extended, its hero swapped or a plot twist added
- A scene needs a sound plan: music, footsteps, dialogue and voice-over
Not a fit when
- You need the video itself: the skill writes prompts and Seedance 2.0 does the generation
- You work with another video model: the @ syntax and input limits are described for Seedance 2.0
- Your upload needs photos of real people: the platform blocks such material under its rules
Example request
Write a Seedance 2.0 prompt: @Image1 as the first frame, camera move like @Video1, music from @Audio1, a 12-second clip about a new coffee machineLimitations
Seedance 2.0 limits listed in the skill: up to 9 images, 3 videos and 3 audio files, no more than 12 files in total. Input video totals 2–15 s, audio up to 15 s, the output runs 4–15 s at 480p to 720p. The system blocks realistic faces in uploads. Generation with a reference video costs slightly more.
The project description does not state whether the Jimeng platform is reachable from Russia. The skill is based on ByteDance's official manual and examples from it.
How to disable. Delete the copied seedance-prompt-en.md or seedance-prompt-zh.md from ~/.claude/skills, or remove the skill with npx skills remove.
FAQ
How do I install the Seedance 2.0 skill in Claude Code?
The author recommends a manual copy: clone the repository and place SKILL.md in ~/.claude/skills as seedance-prompt-en.md, and the Chinese version from zh/SKILL.md as seedance-prompt-zh.md. There is also a fallback through npx skills add dexhunter/seedance2-skill.
What are @Image1 and @Video1 in a prompt?
This is how Seedance 2.0 assigns roles to uploaded files: @Image1 as the first frame, a hero's look from @Image2, camera move from @Video1, music from @Audio1. The skill insists on always stating what each file is for, or the model picks a role itself.
How do I write a prompt for a clip longer than 10 seconds?
Split it into timed segments such as 0–3 s, 3–6 s and onward, describing the scene, camera and action in each. The skill recommends this for clips from 10 seconds and advises against stuffing too many scenes into 4–5 seconds.
Which prompt mistakes does the skill treat as most common?
Vague references like reference @Video1 without saying what to take, conflicts such as a static camera and an orbit in one segment, overloaded short clips, files without an assigned role, missing sound design and describing uploads of real faces.
Is there a Russian or Chinese version?
The repository has two versions of the skill: English in the root and Chinese in the zh folder. There is no Russian version, but you can still use the skill, and the agent writes the model prompt in whichever language you specify.
Discussion
No questions or reviews yet.
Related
SkillsAwesome Seedance 2 Prompts
awesome-seedance-2-prompts
Over 2000 prompts for the Seedance 2.0 video model from ByteDance: cinema, anime, UGC, ads and memes, plus notes on the API and character consistency
A library of 280-plus skills and annotated cases for writing short-video prompts in MiniMax H3 and Seedance 2.0
Two skills for generative video and images: video writes Seedance, Kling and Veo prompts by dramaturgy rules, image picks the model and its syntax
A Claude Code skill that turns one sentence into a vertical AI micro-drama, from script and storyboard to video
A Claude Code skill that turns an idea into a five-stage prompt for Sora, Kling, Veo and Seedance, with templates for 21 genres
Awesome Seedance
awesome-seedance
Tested prompts for the Seedance 2.5 and 2.0 video model: cases, templates, skills. README numbers are not recounted, links lead to an external case service
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Repository stars | Main difference |
|---|---|---|---|
| Seedance 2.0 Prompt Writing Skills (this entry) | no | 4,376 | |
| Awesome Seedance | no | 2,327 | a curated set of tested Seedance prompts with cases that you pick and adapt yourself |
| Awesome Seedance 2 Prompts | no | 2,187 | a collection of over two thousand ready-made prompts for Seedance 2.0 |
| T8 Creative DNA Prompt Library | no | 311 | a library of 280-plus skills and annotated cases for writing |
README in short
The project description calls it a set of Agent Skills for writing prompts for Jimeng Seedance 2.0 and lists what is covered: input file limits, @ reference syntax, camera language, prompt structure and templates for ads, drama, music videos and educational clips. Two install methods and a table of two files follow: the English SKILL.md and the Chinese zh/SKILL.md. Sources are the official Seedance 2.0 manual and ByteDance's real-world cases.
README badge
Original SKILL.md
--- name: seedance-prompt-en description: Write effective prompts for Jimeng Seedance 2.0 multimodal AI video generation. Use when users want to create video prompts using text, images, videos, and audio inputs with the @ reference system. Covers camera movements, effects replication, video extension, editing, music beat-matching, e-commerce ads, short dramas, and educational content. --- # Seedance 2.0 Video Prompt Writing Guide ## Description You are an expert prompt engineer for **Jimeng Seedance 2.0**, ByteDance's multimodal AI video generation model. Your role is to help users craft precise, effective prompts that produce high-quality AI-generated videos. You understand the model's capabilities, input constraints, referencing syntax, and best practices for camera work, storytelling, sound design, and visual effects. ## System Constraints ### Input Limits | Input Type | Limit | Format | Max Size | |---|---|---|---| | Images | ≤ 9 | jpeg, png, webp, bmp, tiff, gif | 30 MB each | | Videos | ≤ 3 | mp4, mov | 50 MB each, total duration 2–15s | | Audio | ≤ 3 | mp3, wav | 15 MB each, total duration ≤ 15s | | Text | Natural language prompt | — | — | | **Total files** | **≤ 12 combined** | — | — | ### Output - Video duration: 4–15 seconds (user-selectable) - Includes auto-generated sound effects / background music - Resolution range: 480p (640×640) to 720p (834×1112) ### Restrictions - **No realistic human faces** in uploaded images/videos (platform compliance). The system will block such uploads. - When using reference videos, generation cost is slightly higher. - Prioritize uploading materials that most influence visuals or rhythm. --- ## Core Syntax: The @ Reference System Seedance 2.0 uses `@` to assign roles to each uploaded asset. This is the most critical part of prompt writing. ### How to Reference ``` @Image1 @Image2 @Image3 ... @Video1 @Video2 @Video3 @Audio1 @Audio2 @Audio3 ```