Skip to content
Agentmart
Submit a tool

giggle-seedance2-gen

Use when the user wants to generate AI video, optimize video prompts, or work with Seedance 2.0.

Skill by giggle-official

IndexedMIT
Write a review

Chalk Labs

Need help deploying giggle-seedance2-gen?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/giggle-official-giggle-seedance2-gen

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Use when the user wants to generate AI video, optimize video prompts, or work with Seedance 2.0. Covers text-to-video, image-to-video, omni multi-modal via giggle.pro API with built-in prompt engineering. Triggers: generate video, AI video, seedance, giggle, 视频提示词, image to video, text to video, omni, 多模态, 短剧, 广告视频, 首帧图, 角色参考. (User may say 即梦 as generic Seedance habit; delivery stays giggle.pro.)

Use cases: backend, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found
Runs as
Python package

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/giggle-seedance2-gen

What's inside

Skill name
giggle-seedance2-gen
Description
Use when the user wants to generate AI video, optimize video prompts, or work with Seedance 2.0. Covers text-to-video, image-to-video, omni multi-modal via giggle.pro API with built-in prompt engineering. Triggers: generate video, AI video, seedance, giggle, 视频提示词, image to video, text to video, omni, 多模态, 短剧, 广告视频, 首帧图, 角色参考. (User may say 即梦 as generic Seedance habit; delivery stays giggle.pro.)
License
MIT
Instructions
321 lines in SKILL.md · view on GitHub

Before installing

  • Ships executable scripts

Files (28)

  • SKILL.md
  • 01-穿越万象-时间穿越.md · prompts/
  • 02-穿越万象-空间穿越.md · prompts/
  • 03-穿越万象-维度穿越.md · prompts/
  • 04-太极生万物.md · prompts/
  • 05-时间扭曲者.md · prompts/
  • 06-一滴水的宇宙.md · prompts/
  • 07-极简爆款试验.md · prompts/
  • 08-基于调研的爆款提示词.md · prompts/
  • 09-延展创意集.md · prompts/
  • 10-一站式测试清单.md · prompts/
  • 11-用户反馈延展.md · prompts/
  • openclaw-《山海战纪》.md · prompts/
  • openclaw-🏙️ 城市恋人.md · prompts/
  • openclaw-🐱 宠物IP短剧全案.md · prompts/
  • openclaw-🤪 10个无厘头反转段子.md · prompts/
  • creative-strategy.md · references/
  • examples.md · references/
  • image-generation.md · references/
  • long-video-strategy.md · references/
  • platform-specs.md · references/
  • production-pipeline.md · references/
  • prompt-engineering.md · references/
  • seedance-prompt-skill-legacy-SKILL.md · references/
  • vocabulary.md · references/
  • requirements.txt
  • generation_api.py · scripts/
  • requirements.txt · scripts/

SKILL.md (first lines)

# Giggle · Seedance 2.0 视频生成与提示词(合并版)

**Source**: [giggle-official/skills](https://github.com/giggle-official/skills) · **API:** [giggle.pro](https://giggle.pro/) · **Script:** `scripts/generation_api.py`

本技能 = **Giggle API 成片**(`scripts/generation_api.py`)+ **Seedance 2.0 提示词工程资料库**(`references/`、`prompts/`)。

- **主流程**:优化提示词 → API Key → 用 `text` / `image` / `omni` 提交并轮询结果。
- **深度写法**:创意钩子、分镜模板、长视频流水线、词表与范例见资源索引。

## 仓库内资源索引

| 路径 | 用途 |
|------|------|
| [references/prompt-engineering.md](references/prompt-engineering.md) | **提示词工程规范**:10 大能力、4 种结构模板、交互流程、质量自检 |
| [references/creative-strategy.md](references/creative-strategy.md) | 写什么、≤15s 与长片创意策略 |
| [references/production-pipeline.md](references/production-pipeline.md) | 长视频前期流水线(角色→分镜→生成→拼接) |
| [references/long-video-strategy.md](references/long-video-strategy.md) | 分段、延长、衔接、锁定语句、输出模板 |
| [references/examples.md](references/examples.md) | 场景与多模态示例(10 大能力分类) |
| [references/vocabulary.md](references/vocabulary.md) | 运镜、画质、大气效果词库 |
| [references/image-generation.md](references/image-generation.md) | 角色参考图 / 首帧图前置生图规范 |
| [prompts/](prompts/) | 主题扩展包、OpenClaw 全案短文 |

---

## ⚠️ 平台与 API 约束(以 Giggle 为准)

资料库为通用 Seedance 提示词最佳实践;**视频与多模态成片仅通过 giggle.pro 提交**。实际调用本仓库脚本时遵守下表:

| 项目 | Giggle(本脚本) |
|------|------------------|
| 语言 | 与**用户输入语言一致**(见下节 Language Rule);资料里中文模板可套用结构 |
| 时长 | **4–15** 秒整数,`--duration` |
| 画幅 / 清晰度 | `--aspect-ratio`、`--resolution` |
| omni 图片 | 最多 **9** 张,`url:` 或 `base64:` |
| 参考音 / 视频 | 仅 `url:`,`--audios`、`--videos` |
| 提示词里的 `@图片N` / `@视频N` | **仅编号习惯**;必须通过 CLI 传入**对应真实 URL**,且**全部**引用都要传 |

若资料写「混合素材 ≤12 个」与 Giggle 限制冲突,**取更严一侧**(尤其图片 ≤9)。

### `@引用` → Giggle CLI(omni)

| 提示词习惯 | 映射到 |
|------------|--------|
| `@图片1` … `@图片N` | `--images "url:..."`(按编号顺序,可多参数或多次 `--images`,脚本会合并) |
| `@视频1` … | `--videos "url:..."` |
| `@音频1` … | `--audios "url:..."` |

---

## ⚠️ Language Rule — MUST FOLLOW

**Match the user's input language exactly. Never translate. Never output a second language version.**

- User writes in Chinese → optimize in Chinese → pass Chinese prompt to API  
- User writes in English → optimize in English → pass English prompt to API  
- Output **ONE** version of the optimized prompt. No bilingual display. No "API submission version".

(资料库中大量中文范例:仅借鉴**结构与章法**,输出语言仍以上述规则为准。)

---

## Step 1: 提示词优化(必须完成后才能进入 Step 2)

用户输入往往描述不准确。**必须先走完以下完整优化流程,生成高质量 prompt,用户确认后再提交 API。**

完整规范见 [references/prompt-engineering.md](references/prompt-engineering.md),以下为 SKILL.md 内联的执行摘要。

---

### Seedance 2.0 核心能力速查

| # | 能力 | 提示词核心模式 |
|---|------|---------------|
| 1 | **一致性控制** | `[角色]@图片N + [动作/剧情] + [场景]@图片N` |
| 2 | **运镜/动作复刻** | `参考@视频1的[运镜/动作/节奏] + [主体]@图片N` |
| 3 | **创意/特效复刻** | `参考@视频1的[特效/转场] + 将[元素]替换为@图片N` |
| 4 | **剧情补全** | `[分镜脚本] + [演绎方式] + [音效/台词]` |
| 5 | **视频延长** | `将@视频1延长Xs + [新增内容]` |
| 6 | **声音控制** | `[画面] + 音色参考@视频1 + "台词"` |
| 7 | **一镜到底** | `一镜到底 + @图片1@图片2... + 全程不切镜头` |
| 8 | **视频编辑** | `将@视频1中的[A]换成@图片1 + [修改说明]` |
| 9 | **音乐卡点** | `@图片1...@图片N + 参考@视频1的画面节奏/卡点` |
| 10 | **情绪演绎** | `[角色] + [情绪变化描述] + [运镜配合]` |

各能力详细示例见 [references/examples.md](references/examples.md)。

---

### Phase 0:判断内容类型(最关键的一步)

| 场景 | 路径 | 核心任务 |
|------|------|---------|
| ≤15秒 | 路径A | 帮用户找到一个**有传播力的单一视觉 hook**;参考 [creative-strategy.md](references/creative-strategy.md) 爆款模式库 |
| >15秒 | 路径B | 走完整前期流水线:角色设计 → 角色卡图提示词 → 分镜脚本 → 首帧图提示词 → 逐镜头视频提示词;参考 [production-pipeline.md](references/production-pipeline.md) |

### Phase 1:获取用户创意

用户描述想要生成的内容。

### Phase 2:确认关键参数(已明确可跳过)

1. **时长**:短片 4–8s / 中 9–12s / 长 13–15s / 超长 >15s
2. **比例**:横 16:9 / 竖 9:16 / 方 1:1
3. **素材**:纯文本 / 有图片 / 有图+视频 / 全模态
4. **补充偏好**(可选):情绪氛围、镜头风格、用途场景

### Phase 3:生成优化提示词(≤15秒输出 2–3 个版本,>15秒按流水线输出)

**提示词结构速查:**

| 场景 | 推荐结构 |
|------|---------|
| ≤12秒 | `[风格总纲],[主体],[动作],[环境/光影],[运镜],[音效]` |
| 13–15秒 | 时间戳分镜法(0-3s / 4-8s / 9-12s / 13-15s 逐段描述) |
| 短剧对白 | 画面 + 台词(角色/情绪标注)+ 音效分离 |
| 史诗大制作 | 品质锚定 + 大气连贯声明 + 逐段 + 光影三层 + 收束句 |

**运镜三级速查:**
- **基础动作**:Pan / Dolly / Zoom / Crane / Orbit / Tracking

From giggle-official/skills, MIT.

More from giggle-official/skills

Signals

Stars
0
Forks
2
Open issues
0
Last commit (this folder)
Jul 3, 2026 (4 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from giggle-official

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get giggle-seedance2-gen running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying giggle-seedance2-gen?

CallGet help
Cite this page

giggle-seedance2-gen skill. Agent Mart. https://agentmart.dev/skills/giggle-official-giggle-seedance2-gen (last updated 2026-10-24)