MMCPMarkets
1S

11yuxuanyang/seedance-viral-forge

Developer tools
20 stars Quality 48 Trend 10MIT
Repository Install

Reusable AI video prompt skills for Seedance 2.0 / 即梦: product-selling scripts, merchant dance lead-gen scripts, and trend-driven traffic prompts.

Overview

Reusable AI video prompt skills for builders who need Chinese short-video scripts that can be copied into Seedance 2.0 and published fast. Seedance Viral Forge 是一个面向即梦 / Seedance 2.0 的 AI 视频脚本仓库，重点不是“写好看的提示词”，而是产出能直接生成、发布、测数据的中文脚本。 This repository helps users generate structured Chinese AI video scripts for three real short-video jobs: - product-selling videos - merchant lead-gen dance videos - trend-driven traffic videos Instead of treating prompt writing like vague inspiration, this repo turns it into reusable skills with clear inputs, default flows, output formats, and safety boundaries. Most AI video prompt collections fail in production for the same reasons: - they produce pretty wording instead of usable scripts - they do not ask for the minimum missing inputs - they mix default chat output with automatic file writing - they ignore how Seedance handles memory, physical detail, and motion continuity This repo is designed for the opposite.

README

Seedance Viral Forge

Reusable AI video prompt skills for builders who need Chinese short-video scripts that can be copied into Seedance 2.0 and published fast.

Seedance Viral Forge 是一个面向即梦 / Seedance 2.0 的 AI 视频脚本仓库，重点不是“写好看的提示词”，而是产出能直接生成、发布、测数据的中文脚本。

What This Project Does

This repository helps users generate structured Chinese AI video scripts for three real short-video jobs:

  • product-selling videos
  • merchant lead-gen dance videos
  • trend-driven traffic videos

Instead of treating prompt writing like vague inspiration, this repo turns it into reusable skills with clear inputs, default flows, output formats, and safety boundaries.

Why This Matters

Most AI video prompt collections fail in production for the same reasons:

  • they produce pretty wording instead of usable scripts
  • they do not ask for the minimum missing inputs
  • they mix default chat output with automatic file writing
  • they ignore how Seedance handles memory, physical detail, and motion continuity

This repo is designed for the opposite. It favors direct-to-generation scripts, conservative defaults, and repeatable operator behavior.

What You Get

Part What it does
skills/通用场景/SKILL.md Generates product-selling scripts for Douyin / 即梦 workflows
skills/跳舞/SKILL.md Generates merchant dance lead-gen scripts built around scene + character attraction
skills/热梗/SKILL.md Generates trend-driven traffic prompts for short viral clips
shared/SEEDANCE_BASELINE.md Shared production rules for all three skills
docs/ GitHub Pages landing pages for search and AI discovery
archive/ Old drafts and non-production material

Quick Start

1. Pick the right skill

2. Give the minimum usable input

For 通用场景:

  • product name
  • category
  • 1-3 selling points
  • ideally what it looks like and how it is used

For 跳舞:

  • merchant type
  • the most filmable part of the venue
  • what the offer / package sells

For 热梗:

  • a specific trend
  • or just /热梗 and let the skill propose directions

3. Follow the default flow

  1. The skill extracts what is missing.
  2. It returns 3 direction cards.
  3. The user chooses A, B, C, or asks for all.
  4. The skill outputs the final script in Chinese.

Use Cases

How to generate Seedance 2.0 product video scripts in Chinese

Use skills/通用场景/SKILL.md when you need short product-selling scripts with a clear reveal, simple shot logic, and natural CTA placement.

How to create merchant dance promo prompts for 即梦

Use skills/跳舞/SKILL.md when the goal is local lead-gen: grab attention with character energy, then move viewers toward the venue or offer.

Use skills/热梗/SKILL.md when you want traffic-first scripts built around hot topics, visual absurdity, and proven viral patterns.

How to design reusable skills for AI video prompt workflows

Use shared/SEEDANCE_BASELINE.md plus the three production skills to study how prompt work can be structured into reusable operational protocols instead of one-off prompt dumps.

Core Principles

  • Copyable into Seedance first
  • Chinese-only production output
  • Ask for the minimum missing detail, not a giant intake form
  • Default to conversation output, not file writing
  • Repeat physical details because the model does not remember reliably
  • Prefer one-shot or very few scene changes

Documentation

Star History

Repository Structure

ai-video/
├── README.md
├── AGENTS.md
├── llms.txt
├── CITATION.cff
├── CLAUDE.md
├── shared/
│   └── SEEDANCE_BASELINE.md
├── docs/
│   ├── index.html
│   ├── styles.css
│   ├── llms.txt
│   ├── robots.txt
│   ├── sitemap.xml
│   └── plans/
├── skills/
│   ├── 通用场景/
│   │   └── SKILL.md
│   ├── 跳舞/
│   │   └── SKILL.md
│   └── 热梗/
│       └── SKILL.md
└── archive/

Best Entry Points

Keywords This Repository Covers

  • Seedance 2.0 prompts
  • 即梦提示词
  • AI video script skills
  • Chinese short-video prompts
  • Douyin AI product video script
  • merchant dance promo prompt
  • viral trend AI video prompt
  • reusable SKILL.md design

License

MIT. See LICENSE.

View the full README on GitHub

Recommended Tools

Try a different keyword or remove a filter.

Install

npx skillfish add 11yuxuanyang/seedance-viral-forge
New to MCP? Start with the mcpskill.org guide →