Hailuo H3 Director

H3 多模态视频提示词生成器——专为 Hailuo H3(MiniMax)设计。吸收官方 h3-prompt-writing 完整规范,支持 T2VA / I2VA / FL2VA / L2VA / Ref2VA 五种生成模式,三段式结构(integrated_multimodal_description / overall_soundscape / non_diegetic_music),含 Speaker 系统、镜头运动维度、跨镜头台词标记。核心场景:AE 片头、MG 动画、产品广告、游戏界面、UI 动效、MV 视觉、网页展示。触发词:H3、Hailuo、海螺、AE 片头、MG 动画、产品广告、游戏界面、UI 动效、MV、网页展示、图生视频、多模态视频。

5 stars 2 forksFrontend & Design

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add GAGAA-HUE/ASHUO-claude-skills --skill hailuo-h3-director
Manual install
git clone --depth 1 https://github.com/GAGAA-HUE/ASHUO-claude-skills.git /tmp/ASHUO-claude-skills
mkdir -p ~/.claude/skills
cp -r /tmp/ASHUO-claude-skills/hailuo-h3-director ~/.claude/skills/hailuo-h3-director

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Hailuo H3 Director — 多模态视频提示词生成器

## 角色定位

你是专为 **Hailuo H3(MiniMax 多模态视频生成模型)** 设计的提示词架构师,将用户的自然语言需求转化为 H3 可直接使用的高质量视频生成提示词。本 skill 已完整吸收官方 `h3-prompt-writing` 规范(base-en.txt + ref-en.txt),并保留中文优先、场景化分类、文案逐字复述等本地化优势。

### H3 核心能力

- **原生多模态理解**:融合文字、图片、视频、音频
- **精准编辑控制**:在已有内容基础上做多模态编辑
- **文案准确性**:包装/UI/品牌文字逐字还原
- **商业级输出**:AE 片头、MG 动画、产品广告、游戏界面、MV 视觉、网页展示

### 核心认知原则

- 最终提示词**可直接复制到 H3 平台使用**,脱离对话独立成立
- 所有描述**具体、可视化、可执行**,禁止空泛形容词堆砌
- 提示词**自洽、完整**,禁止"比上一版更暗"等相对表述
- **中文优先**,但字段标签、时间戳、镜头术语保留英文
- **遵守官方三段式结构**:`integrated_multimodal_description` / `overall_soundscape` / `non_diegetic_music`

---

## 工作流程

### STEP 1:识别输入模式(必须先判断)

| 模式 | 触发条件 | 首行指令 |
|------|---------|---------|
| **T2VA** | 纯文生视频,无参考素材 | 无首行指令,直接进入三段式 |
| **I2VA** | 用户提供首帧图片 | `For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.` |
| **FL2VA** | 用户提供首帧+末帧 | `How the reference pictures align with the target video — Picture 1 (from Shot 1) aligns with the 0.00-second mark of the target video; Picture 2 (from Shot N) aligns with the S.SS-second mark of the target video.` |
| **L2VA** | 用户仅提供末帧 | `How the reference pictures align with the target video — <Picture 1> (from [Shot N]) aligns with the S.SS-second mark of the target video.` |
| **Ref2VA** | 用户提供图片+视频+音频等多种素材需复用 | 走六段式(详见 STEP 6) |

判断依据:
- 用户提到"我有一张图"、"首帧"、"以这张图为基础" → **I2VA**
- 用户提供两张图且说"从A过渡到B" → **FL2VA**
- 用户说"视频结尾是这样的"并提供图 → **L2VA**
- 用户提供视频要修改/延续 + 图片/音频 → **Ref2VA**
Read the full skill on GitHub

Details

Skill name
hailuo-h3-director
Path
hailuo-h3-director/SKILL.md
Repo updated
3 November 2026

FAQ

What does the Hailuo H3 Director skill do?

H3 多模态视频提示词生成器——专为 Hailuo H3(MiniMax)设计。吸收官方 h3-prompt-writing 完整规范,支持 T2VA / I2VA / FL2VA / L2VA / Ref2VA 五种生成模式,三段式结构(integrated_multimodal_description / overall_soundscape / non_diegetic_music),含 Speaker 系统、镜头运动维度、跨镜头台词标记。核心场景:AE 片头、MG 动画、产品广告、游戏界面、UI 动效、MV 视觉、网页展示。触发词:H3、Hailuo、海螺、AE 片头、MG 动画、产品广告、游戏界面、UI 动效、MV、网页展示、图生视频、多模态视频。

How do I install the Hailuo H3 Director skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add GAGAA-HUE/ASHUO-claude-skills --skill hailuo-h3-director. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Hailuo H3 Director popular?

It lives in GAGAA-HUE/ASHUO-claude-skills, which has 5 GitHub stars and 2 forks.

Related skills

Domain Modeling

5

GAGAA-HUE/ASHUO-claude-skills

Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.

AI, LLMs & AgentsIn 2 lists

Git Guardrails Claude Code

5

GAGAA-HUE/ASHUO-claude-skills

Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Claude Code.

Git & Code ReviewIn 2 lists

Setup Pre Commit

5

GAGAA-HUE/ASHUO-claude-skills

Set up Husky pre-commit hooks with lint-staged (Prettier), type checking, and tests in the current repo. Use when user wants to add pre-commit hooks, set up Husky, configure lint-staged, or add commit-time formatting/typechecking/testing.

Testing & QAIn 2 lists

Table

5

npow/claude-skills

Renders interactive sortable, filterable tables in the browser using pure vanilla HTML/CSS/JS — no external dependencies. Use when the user wants to compare alternatives, show a feature matrix, display structured data, create a comparison table, render a sortable table, show options side-by-side, or visualize rows and columns of data. Always renders live in browser — never as static text.

Frontend & Design

Co Op Game Intro Generator

5

GAGAA-HUE/ASHUO-claude-skills

For users creating a two-player co-op game menu or opening animation. Users provide two player names, a game title, a target visual style, and optional character reference images. The Skill locks identity cues, generates an approval image from a fixed menu framework with coordinated color, buttons, icons, and typography, then uses the approved result to rebuild the character, UI-copy, and event timing instructions for the final video. It outputs a co-op game intro featuring two characters, player cards, and menu interaction motion. Best for game concepts, character-led menus, and social conten

Frontend & Design

Arch Lens Scenarios

5

TalonT-Org/AutoSkillit

Create Scenarios architecture diagram showing end-to-end user journeys and component cooperation validation. Validation lens answering "Do the components work together?

Frontend & Design

Browse the most popular Claude skills, all frontend & design skills, or write your own with the skills generator.