DOCSEARCHbyAlgolia
Submit Docs

Seedance 2.0 Prompt Writing Skills

repository·main·Indexed 25 days ago

https://github.com/dexhunter/seedance2-skill

Specialized prompt engineering skills for Jimeng Seedance 2.0, ByteDance's multimodal AI video generation model. Designed for AI agents like Claude Code, Cursor, and Cline, this library provides guidance on input constraints, @ reference syntax for assets, camera language, time-segmented prompt structures, and templates for ads, dramas, and music videos.

Tokens
2.7K
Snippets
2
Records
13
Agent score
86%

What's inside seedance2-skill

  1. Overview of Seedance 2.0 Prompt Writing Skills

    main

    These Agent Skills are designed for the Jimeng Seedance 2.0 multimodal AI video generation model. They provide AI agents (such as Claude Code, Cursor, or Cline) with specialized knowledge to write effective video prompts.

    Key capabilities include:

    • Input constraints and @ citation syntax.
    • Camera movement language and prompt structural patterns.
    • Ready-to-use templates for scenarios like advertising, short dramas, music videos (MV), and science education.

  3. Extend or Edit Videos with Seedance 2.0

    main

    Video Extension

    Use the Extend command to add duration to an existing video.

    Forward Extension:

    Extend @Video1 by 15 seconds.
1–5s: [description]
6–10s: [description]
11–15s: [description]

    Note: Set the generation duration to match the extension length (e.g., extend 5s → select 5s generation).

    Backward Extension (Prepending):

    Extend backward 10s. In warm afternoon light, the camera starts from...

    Video Editing (Modification)

    To change specific elements while preserving the rest of the video, use a 'subvert' or 'modify' approach:

    Subvert @Video1's plot — the man's expression shifts from tenderness to icy cruelty. In an unguarded moment, he shoves the female lead off the bridge...

  4. Install Seedance 2.0 Prompt Writing Skills manually

    main

    To install the skills manually, clone or download this repository and copy the skill files into your Claude skills directory. You can choose between the Chinese version or the English version.

    1. Create the directory: mkdir -p ~/.claude/skills
    2. Copy the Chinese version: cp zh/SKILL.md ~/.claude/skills/seedance-prompt-zh.md
    3. Copy the English version: cp SKILL.md ~/.claude/skills/seedance-prompt-en.md
    mkdir -p ~/.claude/skills

# 中文版
cp zh/SKILL.md ~/.claude/skills/seedance-prompt-zh.md

# 英文版
cp SKILL.md ~/.claude/skills/seedance-prompt-en.md

  5. Understand Jimeng Seedance 2.0 Input Constraints

    main

    Before writing prompts, ensure your assets comply with the following system limits for Jimeng Seedance 2.0:

    Input TypeLimitFormatMax Size
    Images≤ 9jpeg, png, webp, bmp, tiff, gif30 MB each
    Videos≤ 3mp4, mov50 MB each, total duration 2–15s
    Audio≤ 3mp3, wav15 MB each, total duration ≤ 15s
    TextNatural language prompt——
    Total files≤ 12 combined——

    Output Specifications:

    • Video duration: 4–15 seconds (user-selectable).
    • Resolution: 480p (640×640) to 720p (834×1112).
    • Audio: Includes auto-generated sound effects or background music.

    Important Restrictions:

    • No realistic human faces in uploaded images/videos (platform compliance will block these).
    • Using reference videos results in slightly higher generation costs.

  6. Follow the Seedance 2.0 Prompt Creation Workflow

    main

    When creating a Seedance 2.0 prompt, follow these steps to ensure high-quality video generation:

    1. Clarify the goal: Determine the video type (e.g., Ad, drama, MV, educational, vlog).
    2. Identify available assets: List all provided images, videos, and audio files.
    3. Assign roles: Map each asset to a specific function (e.g., first frame, character ref, camera ref).
    4. Structure the prompt:
      • Start with subject and scene setup.
      • For videos > 8s, add time-segmented action descriptions.
      • Specify camera movements.
      • Add audio/sound design instructions.
      • Include style modifiers.
    5. Check constraints: Ensure total files $\le$ 12, no real human faces are described, and durations are within limits.
    6. Optimize: Remove ambiguity and ensure every @reference has a clear, assigned role.

  7. Use Seedance 2.0 Prompt Writing Skills with AI agents

    main
    Once installed, you can use these skills with compatible agents like Claude Code, Cursor, Cline, and others. Simply ask your AI agent to help you write a Seedance 2.0 video prompt. The skills provide guidance on input constraints, @ reference syntax, camera language, prompt structure patterns, and templates for various use cases (ads, dramas, MVs, etc.).

  8. Structure a Seedance 2.0 Prompt

    main

    To achieve high-quality results, follow the recommended prompt structure blueprint.

    Standard Formula

    Combine these elements in order: [Subject/Character Setup] + [Scene/Environment] + [Action/Motion Description] + [Camera Movement] + [Timing Breakdown] + [Transitions/Effects] + [Audio/Sound Design] + [Style/Mood]

    Time-Segmented Prompts (Recommended for 10s+ videos)

    For precise control over longer videos, break the prompt into timed segments:

    0–3s: [opening scene description, camera, action]
3–6s: [mid-section development]
6–10s: [climax or key action]
10–15s: [resolution, ending shot, final text/branding]

  9. Avoid common Seedance 2.0 prompt mistakes

    main

    To prevent generation errors or poor quality, avoid these common mistakes:

    • Vague references: Do not just say reference @Video1. Specify the purpose (e.g., camera, action, effects, or rhythm).
    • Conflicting instructions: Avoid requesting contradictory movements (e.g., asking for both "static camera" and "orbit shot") in the same segment.
    • Overloading: Do not pack too many scenes into a short duration (e.g., 4–5 seconds); keep actions physically plausible.
    • Missing @ assignments: Ensure every uploaded file is referenced with a clear purpose using the @ syntax.
    • Ignoring audio: Always include audio direction to improve output quality.
    • Forgetting duration: Match prompt complexity to the selected generation length.
    • Real faces: Do not describe or attempt to upload real human photos; the system will block these.

  10. Use Seedance 2.0 prompt templates

    main

    Use these templates as starting points for different video types:

    Product Ad (15s)

    Reference @Video1's editing style and camera transitions. Replace @Video1's
product with @Image1 as the hero product. Create a 15-second product
showcase video.
0–3s: Product enters frame with dynamic rotation, close-up on surface
texture and logo details.
4–8s: Multiple angle transitions — front, side, back — with product
highlight scanning light effects.
9–12s: Product in lifestyle context showing usage scenario.
13–15s: Hero shot with brand tagline appearing, background music builds
to resolution.
Sound: Reference @Video1's background music. Add product interaction
sound effects.

    Short Drama (15s)

    Scene (0–5s): Close-up on the character's reddened eyes, finger pointing
accusingly, tears streaming down. Emotion on the edge of collapse.
Dialogue 1 (Character A, choking with rage): "What exactly are you trying
to take from me?"
Scene (6–10s): The other character trembles, holding up evidence,
red-eyed, stepping forward. Camera sweeps past background details.
Dialogue 2 (Character B, urgent and choked): "I'm not deceiving you! This is what he entrusted to me!"
Scene (11–15s): Evidence is revealed, Character A freezes — expression
shifts from anger to shock, hands slowly rise.
Sound: Urgent piano + static interference, sobbing, button click sound,
ending with a muffled voice blending in.
Duration: Precise 15 seconds, every frame tight, no filler.

    Dance Video (13s)

    Have the character in @Image1 replicate the dance moves and beat-synced
music from @Video1. Generate a 13-second video. Movements should be
smooth with no stuttering or freezing.

    Scenery Montage (15s)

    @Image1 @Image2 @Image3 @Image4 @Image5 @Image6 — landscape scene
images. Reference @Video1's visual rhythm, inter-scene transitions,
visual style, and music tempo for beat-synced editing.

  11. Use the @ Reference System for Assets

    main

    Seedance 2.0 uses the @ symbol to assign specific roles to uploaded assets. This is critical for controlling how the model uses your images, videos, and audio.

    Reference Syntax

    Use @ followed by the asset identifier:

    @Image1    @Image2    @Image3
@Video1    @Video2    @Video3
@Audio1    @Audio2    @Audio3

    Assigning Roles

    You must explicitly state the purpose of each reference. Common patterns include:

    PurposeExample Syntax
    First frame@Image1 as the first frame
    Last frame@Image2 as the last frame
    Character appearance@Image1's character as the subject
    Scene/backgroundscene references @Image3
    Camera movementreference @Video1's camera movement
    Action/motionreference @Video1's action choreography
    Visual effectscompletely reference @Video1's effects and transitions
    Rhythm/tempovideo rhythm references @Video1
    Voice/tonenarration voice references @Video1
    Background musicBGM references @Audio1
    Sound effectssound effects reference @Video3's audio
    Outfit/clothingwearing the outfit from @Image2
    Product appearanceproduct details reference @Image3

    Multi-Reference Example

    You can combine multiple references in one prompt:

    @Image1's character as the subject, reference @Video1's camera movement and action choreography, BGM references @Audio1, scene references @Image2

  12. Apply Style and Quality Modifiers

    main

    Append these modifiers to the end of your prompts to enhance the visual and auditory output.

    Visual Style

    • Cinematic quality, film grain, shallow depth of field
    • 2.35:1 widescreen, 24fps
    • Ink wash painting style / Anime style / Photorealistic
    • High saturation neon colors, cool-warm contrast
    • 4K medical CGI, semi-transparent visualization

    Mood/Atmosphere

    • Tense and suspenseful / Warm and healing / Epic and grand
    • Comedy with exaggerated expressions
    • Documentary tone, restrained narration

    Audio Direction

    • Background music: grand and majestic
    • Sound effects: footsteps, crowd noise, car sounds
    • Voice tone reference @Video1
    • Beat-synced transitions matching music rhythm