seedance-20

Generate cinematic AI videos via text, image, video, and reference inputs.

6.3k|941|Updated Feb 25, 2026
One-click install
npx skills add https://github.com/Emily2040/seedance-2.0 --skill seedance-20
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: seedance-20
Source: https://github.com/Emily2040/seedance-2.0/tree/main
Command: npx skills add https://github.com/Emily2040/seedance-2.0 --skill seedance-20

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill empowers users to generate and direct cinematic AI videos, transforming simple ideas into high-quality visual content across multiple modalities.

Core Features & Use Cases

  • Quad-Modal Generation: Supports Text-to-Video, Image-to-Video, Video-to-Video, and Reference-to-Video workflows.
  • Director's Toolkit: Provides granular control over camera, motion, lighting, characters, VFX, and audio, all guided by an intent-first philosophy.
  • Use Case: A filmmaker can use this Skill to generate a 10-second dramatic scene where a character walks down a rainy street at night, using a reference image for the character and a specific camera move, all while adhering to a strict 50-word prompt limit.

Quick Start

Use the seedance-20 skill to generate a 10-second video of a character walking down a rainy street at night.

Frequently Asked Questions about seedance-20

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate cinematic AI video with consistent character motion and camera control?

Cinematic AI video generation uses an intent-first prompting system to manage camera movement, motion choreography, and character consistency. You direct the scene by specifying these elements within a 30-100 word prompt target.

What is the best way to convert an existing image into a video with specific lighting and VFX?

Image-to-video conversion is handled through a quad-modal framework, allowing you to integrate lighting design, VFX, and audio synchronization. You use @reference tags for enhanced control over the visual transformation.

Can I use reference images to maintain character consistency across different AI video generations?

Yes, reference-to-video generation maintains character consistency by utilizing @reference tags within the prompt. This ensures character identity remains uniform across different scenes and video outputs.

How does the intent-first prompting system work for directing AI video?

The intent-first prompting system works by prioritizing your cinematic goals—like camera moves, lighting, and VFX integration—within a 30-100 word prompt target. This philosophy ensures the AI focuses on your directorial intent.

Does this AI video generation approach support video-to-video workflows for existing footage?

Yes, video-to-video workflows are fully supported as part of the quad-modal generation framework. This allows you to transform existing footage by applying new camera movements, motion choreography, and lighting designs.

What are the prompt length limitations when generating a cinematic AI video scene?

Cinematic AI video scenes require a strict 30-100 word prompt target to ensure optimal generation. This limitation forces concise, intent-first descriptions of camera, character, and lighting actions.