4.3K ·GitHub
What is this?
This is a set of agent skill files that teach AI coding agents like Claude Code how to write effective prompts for generating videos with ByteDance's Seedance 2.0 model.
This is a set of agent skill files that teach AI coding agents like Claude Code how to write effective prompts for generating videos with ByteDance's Seedance 2.0 model.Is this a video generator, or just instructions for an AI agent?It's not a video generator itself. It's a SKILL.md knowledge file that an AI agent reads so it knows the input constraints, @ reference syntax, camera language, and prompt patterns Seedance 2.0 expects.
It's not a video generator itself. It's a SKILL.md knowledge file that an AI agent reads so it knows the input constraints, @ reference syntax, camera language, and prompt patterns Seedance 2.0 expects.How do I install this skill?Clone or download the repo and manually copy the skill file(s) into your Claude skills directory, or install it via the skills CLI, then just ask your agent to help write a Seedance 2.0 prompt.
Clone or download the repo and manually copy the skill file(s) into your Claude skills directory, or install it via the skills CLI, then just ask your agent to help write a Seedance 2.0 prompt.Which AI agents does it work with?It's designed for Claude Code, Cursor, Cline, and other agents that support skill files; the skill only drafts prompts, so you also need separate access to Seedance 2.0 to actually render video.
It's designed for Claude Code, Cursor, Cline, and other agents that support skill files; the skill only drafts prompts, so you also need separate access to Seedance 2.0 to actually render video.Does this skill generate the video for me?No, it only writes the prompt text. You still need to feed that prompt into Seedance 2.0 yourself to get an actual video output.
No, it only writes the prompt text. You still need to feed that prompt into Seedance 2.0 yourself to get an actual video output.What is Seedance 2.0?It is ByteDance's multimodal AI video generation model (used in the Jimeng app), which this skill teaches agents how to write prompts for.
It is ByteDance's multimodal AI video generation model (used in the Jimeng app), which this skill teaches agents how to write prompts for.What's actually included in the skill files?Two SKILL.md files (English and Chinese) covering input constraints, the @ reference syntax, camera language, prompt structure patterns, and ready-to-use templates for ads, dramas, MVs, and educational content.
Two SKILL.md files (English and Chinese) covering input constraints, the @ reference syntax, camera language, prompt structure patterns, and ready-to-use templates for ads, dramas, MVs, and educational content.Is there a Chinese-language version of the guide?Yes, zh/SKILL.md provides the same prompt-writing guide fully translated into Chinese alongside the English SKILL.md.
Yes, zh/SKILL.md provides the same prompt-writing guide fully translated into Chinese alongside the English SKILL.md.Where do the prompt-writing rules come from?They're based on official ByteDance documentation: the Seedance 2.0 User Manual and the Seedance 2.0 Real-world Cases covering drama production, e-commerce ads, dance imitation, science education, and AI music videos.
They're based on official ByteDance documentation: the Seedance 2.0 User Manual and the Seedance 2.0 Real-world Cases covering drama production, e-commerce ads, dance imitation, science education, and AI music videos.What are similar projects?h3-storyboard-skill, kid-papercraft, reference-video-director, God-minmax-H3, seedance-2-5-video-director, awesome-seedance, bigpeng-hot-gzh, seedance-2.0
h3-storyboard-skill, kid-papercraft, reference-video-director, God-minmax-H3, seedance-2-5-video-director, awesome-seedance, bigpeng-hot-gzh, seedance-2.0Which topics is it in?seedance, ai-video-generation, templates, claude-skill, prompt, text-to-video, chinese-language, multimodal, video-generation, prompt-engineering, skill, cursor
seedance, ai-video-generation, templates, claude-skill, prompt, text-to-video, chinese-language, multimodal, video-generation, prompt-engineering, skill, cursor