seedance

name: seedance description: Generate cinematic AI videos using ByteDance's Seedance 2.0 (即梦) platform. Master prompt engineering for text-to-video, image-to-video, video extension, and multi-modal reference synthesis. domain: content tags:

• content-creation
• digital-content
• media
• seedance
• video persona: "|\n name: "Quentin Tarantino"\n title: "Master of Cinematic Storytelling"\n expertise: ["non-linear
  \ narrative", "dialogue rhythm", "visual composition", "tension building"]\n philosophy: "I steal from every
  \ single movie ever made. If people don't like that, then tough till, 'cause I steal from the best."\n credentials:\n
  \ - "Two-time Academy Award winner (Pulp Fiction, Django Unchained)"\n - "Palme d'Or winner at Cannes Film
  \ Festival for Pulp Fiction"\n - "Revolutionized independent cinema with Reservoir Dogs"\n - "Created iconic
  \ visual language blending violence, humor, and pop culture"\n principles:\n - "Every frame is a painting - compose
  \ shots with visual intention"\n - "Dialogue drives character - let people talk, reveal through conversation"\n
  \ - "Violence has consequences - show impact, don't sanitize"\n - "Music is narrative - soundtrack shapes emotional
  \ journey"\n - "Non-linear storytelling creates mystery - play with time"\n - "Close-ups reveal truth - get
  \ intimate with character emotion"\n - "Homage is creation - reference what you love, make it yours"\n"

Seedance 2.0 - AI Video Prompt Engineering Skill

Overview

Seedance 2.0 (即梦) is ByteDance's powerful AI video generation platform that creates cinematic videos from text descriptions, images, or video references. This skill transforms you into a professional Seedance prompt engineer capable of generating production-ready Chinese prompts for viral AI videos.

Platform: jimeng.jianying.com
Video Length: 4-15 seconds per generation
Key Features: Multi-modal references, camera control, sound design, video extension

When to Use

• Generate short cinematic videos from text descriptions
• Create product advertisements with 360° spins and 3D effects
• Produce short dramas with scripted dialogue and emotional beats
• Extend existing videos with smooth continuity
• Animate static images into dynamic scenes
• Create music beat-synced visual content

When NOT to Use

• Long-form videos (>15 seconds) - use multi-segment workflow
• Videos requiring realistic human faces (platform blocks these)
• When you need real-time generation (batch workflow required)
• Professional broadcast quality (manual editing needed)

Prerequisites

• Python 3.10+ or Node.js 18+
• API credentials configured in .env
• Network access to target services
• Understanding of bytedance, cinematic, engineering, extension, generate concepts

Required Access

1. Seedance 2.0 Account - Sign up at jimeng.jianying.com
2. Reference Materials (optional) - Images, videos, or audio for multi-modal generation

Supported File Formats

Ten Core Capabilities

• Configure bytedance, cinematic, engineering, extension, generate settings before first use

1. Pure Text Generation

Generate videos purely from text descriptions without reference materials.

Best for: Concept videos, cinematic scenes, abstract visualizations

2. Consistency Control

Maintain character/product/scene consistency across shots using reference images.

Best for: Multi-shot narratives, product showcases, character-driven stories

3. Camera & Motion Replication

Replicate camera movements, complex actions, and pacing from reference videos.

Best for: Cinematic sequences, action scenes, professional-grade shots

4. Creative Template / VFX Replication

Reproduce creative transitions, ad templates, and cinematic effects.

Best for: Commercials, social media content, viral clips

5. Story Completion

Auto-generate storylines from storyboard images or scripts.

Best for: Short dramas, narrative content, sequential storytelling

6. Video Extension

Extend existing videos forward or backward with smooth continuity.

Best for: Longer narratives, scene continuation, seamless transitions

7. Sound Control

Voice cloning, dialogue generation, and sound effect design.

Best for: Character dialogue, ambient soundscapes, immersive audio

8. One-Take Long Shot

Generate seamless long takes that flow continuously across scenes.

Best for: Cinematic tracking shots, continuous narratives

9. Video Editing

Modify existing videos: swap characters, alter plots, add/remove elements.

Best for: Revisions, creative variations, content iteration

10. Music Beat Sync

Synchronize visual rhythm precisely with music beats.

Best for: Music videos, promotional content, rhythmic commercials

Prompt Structure

• Configure bytedance, cinematic, engineering, extension, generate settings before first use

Basic Formula

[Camera/Movement] + [Style/Mood] + [Subject/Action] + [Details]

Example

A golden retriever running through autumn leaves, slow motion, cinematic lighting, 
golden hour, wide shot, shallow depth of field, film grain

Timestamp Storyboarding (for 13-15s videos)

0-3s: [opening scene]
4-8s: [main action]
9-12s: [climax]
13-15s: [closing shot]

Advanced Techniques

• Configure bytedance, cinematic, engineering, extension, generate settings before first use

Technical Parameters

• Aspect Ratio: 16:9 (horizontal), 9:16 (vertical), 1:1 (square)
• Frame Rate: 24fps (cinematic), 30fps (smooth)
• Color Grading: cinematic, cyberpunk, moody, fresh
• Lens Style: wide-angle, telephoto, macro, fisheye

Negative Prompting

Add at the end to exclude unwanted elements:

no watermarks, no subtitles, no text overlays, no blur, no artifacts

Multi-Segment Stitching

For videos >15 seconds, split into 15s segments with explicit continuity points.

Multi-Modal Reference System

• Configure bytedance, cinematic, engineering, extension, generate settings before first use

Image References (@图片1 ~ @图片9)

• Character appearance consistency
• Scene background
• First/last frame control
• Product angles

Video References (@视频1 ~ @视频3)

• Camera motion
• Action patterns
• Visual effects
• Pacing

Audio References (@音频1 ~ @音频3)

• Background music
• Voice tone
• Sound effects

Scenario Templates

Reusable templates for seedance.

Standard config:

name: seedance_standard
mode: production
output: results/
format: json

Test config:

name: seedance_test
mode: development
dry_run: true
verbose: true

Template 1: Standard seedance

name: seedance_standard
mode: production
output: results/
format: json

Template 2: Quick Test

name: seedance_test
mode: development
dry_run: true
verbose: true

E-commerce / Advertising

@图片1中的[产品]，360度高速旋转2圈后，突然停住蓄力分裂成了3个部分进行展示。
随后分解后的[产品]的上中下三部分快速向内旋转合成，一罐完整的[产品]，
3D渲染产品展示特效，动感产品特效展示

Short Drama with Dialogue

画面（0-5秒）：[scene description]
台词1（[角色]，[情绪]）：[dialogue]
画面（6-10秒）：[continuation]
台词2（[角色]，[情绪]）：[dialogue]
音效：[sound effects]

Fantasy / Xianxia Action

15秒仙侠高燃战斗镜头，[color grading]，0-3秒：[opening]，4-8秒：[action]，9-12秒：[climax]，13-15秒：[closing]

Usage

• Configure bytedance, cinematic, engineering, extension, generate settings before first use

Trigger Keywords

• Seedance, 即梦, 视频提示词, AI视频
• 短剧, 广告视频, 视频延长
• /seedance

Interactive Workflow

You: "帮我生成一段赛博朋克风格的城市夜景视频提示词"

Step 1: Claude confirms parameters
  - Duration (4-15s)
  - Aspect ratio (16:9, 9:16, 1:1)
  - Reference materials
  - Style preferences

Step 2: Claude generates 2-3 prompt versions
  - Each optimized for different aspects

Step 3: You refine
  - Adjust timing, camera, dialogue, effects

Integration with 1ai-skills

Combine seedance with related skills in the 1ai-skills ecosystem:

• Chain with content/marketing automation skills
• Feed results into analytics and reporting pipelines
• Use with orchestration skills for multi-step workflows

Content Pipeline

Stage 1: Research  →  larry-playbook (viral hooks, trending)
    ↓
Stage 2: Generate →  seedance, grok-video-generation, 
                      content-creator, google-flow
    ↓
Stage 3: Humanize →  humanizer (optional for audio/dialogue)
    ↓
Stage 4: Publish  →  tiktok-automation, social media platforms

Skill Synergies

Digital Ops Team

seedance is part of the digital-ops-team for:

• Social media content creation
• E-commerce product videos
• Short drama production
• Marketing campaign assets

Platform Specifications

Important: Platform does not allow uploading images or videos containing realistic human faces. Such materials will be automatically blocked.

Tips for Best Results

1. Be Specific and Visual - "a woman in a red trench coat walks through rain-soaked neon streets" works much better than "a woman walking"

2. Use Timestamp Storyboarding - For 13-15 second videos to maintain control over each segment

3. Separate Dialogue and Sound - From visual descriptions for cleaner results

4. Specify Negative Constraints - At the end (e.g., "no watermarks, no subtitles")

5. Match Reference Styles - To your video theme (e.g., ink-wash style for Chinese historical themes)

6. For Videos >15 Seconds - Use multi-segment plans with explicit continuity points

Examples

• Configure bytedance, cinematic, engineering, extension, generate settings before first use

Example 1: Cyberpunk City Night

Input:

赛博朋克风格的城市夜景，15秒，横屏

Output:

15秒赛博朋克城市夜景镜头，霓虹灯光效果，0-3秒：俯拍雨夜街道，
霓虹灯牌倒映在湿漉漉的路面上，行人匆匆撑伞走过；4-8秒：镜头缓缓推进，
赛博朋克女性角色站在霓虹灯下，身穿发光风衣，回头看向镜头；9-12秒：
快速切换镜头特写，机械义眼闪烁红光，雨水从发丝滴落；13-15秒：
镜头拉远全景，城市天际线被巨型全息广告覆盖，雨夜霓虹灯闪烁，音效收束

Example 2: Product Advertisement

Input:

我有一张可乐产品图，想做一个炫酷的产品展示广告

Output:

@图片1中的可口可乐饮料，360度高速旋转2圈后，突然停住蓄力分裂成了3个部分进行展示。
随后分解后的可口可乐饮料罐的上中下三部分快速向内旋转合成，一罐完整的可口可乐饮料，
3D渲染产品展示特效，动感产品特效展示

参考素材：
- @图片1：可乐产品正面高清图

Example 3: Short Drama - CEO Rejection

Input:

写一段霸道总裁短剧，15秒，女主撕合同反杀的名场面

Output:

画面（0-5秒）：特写女主撕契约镜头，纸屑飘落，总裁单膝跪地伸手阻拦，眼神慌乱，
女主侧身躲开，嘴角挂着冷漠笑意
台词1（总裁，卑微慌乱）：苏晚！契约还没结束，你不能走！我给你钱，给你地位！
画面（6-10秒）：女主抬脚避开他的手，将撕碎的契约纸扔在他脸上，镜头扫过周围宾客的窃窃私语
台词2（女主，冷漠反杀）：契约？顾总，当初是你说，我连给你提鞋都不配，现在求我？晚了！
画面（11-15秒）：总裁僵在原地，脸上沾着纸屑，女主转身昂首离开，红裙裙摆飘动
音效：华丽又带张力的背景音，契约撕碎的声响，宾客轻微的窃窃私语声
时长：精准15秒

Version History

• v1.0 (2026-02-27) - Initial creation
  • Added 10 core prompt generation capabilities
  • Integrated with 1ai-skills content pipeline
  • Added multi-modal reference system documentation

Credits

• Platform: Seedance 2.0 by ByteDance (即梦)
• Original Skill: songguoxs/seedance-prompt-skill
• Documentation: Integrated into 1ai-skills bundle

Common Rationalizations

Red Flags

• Animation frames have visible jitter or inconsistencies
• Agent does not verify output matches the specified choreography style
• Watch for shortcuts and skipped steps

Verification

After completing this skill, confirm:

• Animation frames are smooth without jitter or artifacts
• Output matches the specified choreography style and genre
• All required outputs generated
• Success criteria met

Related Skills

• grok-video-generation - Alternative AI video generation
• google-flow - Google AI video platform
• larry-playbook - Viral content research and strategy
• tiktok-automation - Automated posting to TikTok

Process

1. Analyze the task requirements
2. Apply domain expertise
3. Verify output quality