This generates prompts for Kling 3.0, the Chinese video AI model that's strongest at native Chinese understanding and synchronized audio. It adapts between three formats: a four-part formula for short clips, a five-layer structure for narrative scenes with dialogue (where Kling's character-directed audio really shines), and motion-only descriptions for image-to-video. The skill walks you through decision trees, asks for missing details without over-questioning, and enforces Kling-specific rules like adding explicit endpoints to prevent generation freezes and using physical interactions to anchor fantasy scenes. If you're making story-driven videos with Chinese dialogue or need that native audio sync, this is built for that workflow. It's less useful for multi-shot storyboards or quick social clips where audio doesn't matter.

GitHub stars
269
last updated
Jun 1, 2026

Install

$ npx -y skills add cclank/lanshu-awesome-ai-video-kit --skill kling-prompter --agent claude-code

Installs into .claude/skills of the current project.

Files

kling-prompter

Kling 3.0(可灵 3.0,快手)的三套主流写法:4 部分基础 / 5 层进阶 / 图生视频。

何时不用此 skill

三套写法决策树

你的需求是?
├── 5 秒短视频、单一场景        → 写法 1(4 部分基础公式)
├── 带剧情、带音频、多角色      → 写法 2(5 层进阶公式)★ Kling 强项
└── 已有参考图,只需动起来      → 写法 3(图生视频专用)★ Kling 强项

工作流程

步骤 1:判断写法

用户的需求关键词写法
短视频、TikTok、抖音、循环、爆款、变身写法 1(4 部分)
对话、剧情、配音、CTA、广告、多角色写法 2(5 层)
这张图、参考这张照片、让它动起来写法 3(图生视频)

步骤 2:澄清缺失要素(必要时)

如果信息不够,问 2-3 个关键问题(不是全部):

场景必问的 2-3 项
短视频主体(具体外观)、动作(动词+终点)、风格(电影感/动漫/赛博朋克)
5 层剧情角色(年龄+服装)、对话内容、要不要原生音频(音乐/音效/voiceover)
图生视频主体在图中做什么变化(不要再描述图本身)、镜头怎么动

步骤 3:按选定写法填充

写法 1:4 部分基础公式
[Subject - 具体视觉细节] [Action - 精确运动+终点] [Context - 3-5 环境元素] [Style - 镜头+灯光+氛围].

详细参考 methodology/09-kling-公式.md。

写法 2:5 层进阶公式(Kling 的看家本领)

强烈推荐用结构化格式,模型按层解析:

Sponsors

Details

Repository
github.com/cclank/lanshu-awesome-ai-video-kit
Categories
AI & Agent Building, Design & UI/UX
License
MIT
First seen
Jun 11, 2026

More from cclank/lanshu-awesome-ai-video-kit

All 7
1cclank avatarmodel-selectorcclank/lanshu-awesome-ai-video-kit根据用户的视频需求(场景、时长、音频、语言、平台限制、预算、是否需要本地部署/角色一致性等),从 16 个主流 AI 视频模型(12 商业 + 4 开源)中推荐最匹配的 1-3 个,并解释为什么。覆盖商业:Seedance 2.0、HappyHorse 1.0、Kling 3.0、Sora 2、Veo 3.1、**Ge0skills.sh installs
2cclank avatarseedance-promptercclank/lanshu-awesome-ai-video-kit把用户的自然语言视频需求转换为符合 Doubao Seedance 2.0 进阶公式的提示词(8 要素:精准主体+动作细节+场景环境+光影色调+镜头运镜+视觉风格+画质+约束条件)。用于"帮我写一个 Seedance 提示词"、"生成视频提示词"、"做个产品广告视频"、"用 Seedance 生成 XX"等触发场景。如0skills.sh installs
3code-and-sorts avatarcalculatorcode-and-sorts/awesome-copilot-agentsPerforms arbitrary-precision arithmetic calculations including addition, subtraction, multiplication, division, and exponents. Use when the user asks to calcula0skills.sh installs
4codeaholicguy avatarchangelogcodeaholicguy/ai-devkitAI DevKit · Update CHANGELOG.md Unreleased items from git commits since the latest release. Use when users ask to update changelog/release notes from recent com0skills.sh installs
5codeaholicguy avatardev-lifecyclecodeaholicguy/ai-devkitAI DevKit · Structured SDLC workflow with 8 phases — requirements, design review, planning, implementation, testing, and code review. Use when the user wants to0skills.sh installs
6codeaholicguy avatardocument-codecodeaholicguy/ai-devkitAI DevKit · Document a code entry point with structured analysis, dependency mapping, and saved knowledge docs. Use when users ask to document, understand, or m0skills.sh installs