Latest release: v1.0.5Download zip
Capabilities
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
The artifacts consistently describe an AI video-production workflow using HTML rendering, animation adapters, TTS, OCR, captions, and image generation.
Instruction Scope
The workflow is scoped around asking for missing video requirements, routing to a production line, creating video assets, and validating/rendering outputs.
Install Mechanism
The registry has no install spec, but the README gives user-directed setup commands to clone a GitHub repo and run pnpm install; this is purpose-aligned but should be reviewed before execution.
Credentials
Node.js, FFmpeg, pnpm, external AI/TTS/OCR/transcription services, and local render tooling are proportionate for video generation, though not all are declared in registry requirements.
Persistence & Privilege
The skill writes project artifacts such as workspace outputs and .pentovideo state files; a separate contribution workflow can also use GitHub/AWS credentials and public publishing commands when explicitly invoked.
Assessment
Before installing or running production commands, verify the GitHub repository and dependencies. Avoid sending confidential scripts, images, PPTs, or audio to external providers unless approved. Confirm explicitly before any publish, GitHub, or AWS-profile command is run.Verification
Tags
What is PentoVideo? | 这是什么？
PentoVideo is an AI-powered video factory that turns text, images, PPTs, and prompts into professional videos via HyperFrames HTML rendering. Four production lines, one engine. Forked from and built upon the open-source PentoVideo framework.
PentoVideo 是一个 AI 视频工厂，通过 HyperFrames HTML 渲染将主题、图片、PPT、口播稿转化为专业视频。四条生产线，一个引擎。基于开源 PentoVideo 框架构建。
| Line 生产线 | Input 输入 | Output 输出 |
|---|---|---|
| A — Pure CSS | Topic / Script | HTML + CSS animation → MP4 |
| B — Image Gen | Topic + SenseNova | AI images + HTML → MP4 |
| C — PPT | PPTX file | OCR → HTML → MP4 |
| D — Images | Images + Script | OCR/align → HTML → MP4 |
| D-Fast ⚡ | Images + Script (ready) | TTS → HTML → MP4 (4 steps) |
Quick Start | 快速开始
# Clone the skill
git clone https://github.com/zhbcher/pentovideo.git
# Install dependencies
pnpm install
# Start video production
# Agent auto-triggers: §0 gate → route → produce
Fastest Path | 最快路径
Have images + script + style? Jump straight to Line D-Fast:
Images + Script → Edge TTS → Scene split → HTML → Render → MP4
See workflows/line-d-fast.md for the 4-step workflow.
Architecture | 架构
§0 Pre-Flight Gate (7-field check)
│
§1 Routing Decision Tree
│
┌─────────────────────┼─────────────────────┐
▼ ▼ ▼
Line A (CSS) Line B (Gen) Line C/D (PPT/Img)
│ │ │
└─────────────────────┼─────────────────────┘
▼
§2 Design System
§3 Prompt Expansion
§4 Build Timeline
§5 HTML Authoring
§6 Quality Gates (lint/validate)
§7 Preview → Render → Deliver
Key Features | 核心功能
- 🛑 Pre-Flight Gate (§0) — 7-field check before any work starts. Prevents context-losing rework.
- 🚀 Quick Jump — 8 scenario-based entry points. Find your path in one click.
- ⚡ Line D-Fast — Images + script → video in 4 steps. Our fastest production line.
- 🎨 18 Style Presets — Tech-dark, neon-gradient, business-green, and more.
- 🎙️ Edge TTS — Free unlimited Chinese TTS. Inline Python command, no script dependency.
- 🖼️ SenseNova Gen — AI image generation via SenseNova U1 Fast. 11 sizes, 16:9 ready.
- ✅ Quality Gates — lint/validate/inspect/contrast/design/animation-map. 6 checks before render.
- 🌐 Bilingual — Chinese + English throughout. Built for Chinese-speaking developers.
Production Lines | 生产线详情
| Line | Workflow | When to Use |
|---|---|---|
| A | line-a-pure-css.md | Topic → HTML+CSS animation |
| B | line-b-image-gen.md | Topic → SenseNova gen → OCR → HTML |
| C | line-c-ppt.md | PPT → OCR → HTML |
| D | line-d-images.md | Images → OCR → HTML |
| D-Fast ⚡ | line-d-fast.md | Images + Script → 4-step fast track |
Tools & Integration | 工具集成
| Tool | Description | File |
|---|---|---|
| Edge TTS | Free Chinese text-to-speech | tools/edge-tts.md |
| SenseNova Gen | AI image generation | tools/sensenova-image-gen.md |
| OCR Check | Image text extraction QA | tools/ocr-check.md |
| Cover Gen | Video cover/thumbnail | tools/cover-generation.md |
| Build Timeline | Scene timing construction | tools/build-timeline.md |
Animations & Transitions | 动画与转场
Full animation adapter support: GSAP, WAAPI, CSS, Anime.js, Three.js, Lottie. CSS + WebGL transition system.
| Adapter | Coverage |
|---|---|
| GSAP | Full timeline, easing, stagger |
| CSS Animations | Keyframes, delays, fill modes |
| WAAPI | element.animate(), deterministic seeking |
| Anime.js | Timelines, seek-driven rendering |
| Three.js | WebGL scenes, camera motion |
| Lottie | JSON + dotLottie, paused playback |
Project Structure | 项目结构
PentoVideo/
├── SKILL.md # Skill definition (bilingual)
├── README.md # This file
├── workflows/ # 4 production lines + line-d-fast
├── tools/ # TTS, image gen, OCR, cover, timeline
├── references/ # Design system, transitions, captions
├── styles/ # 18 style presets
├── animations/ # 6 animation adapters
├── packages/ # CLI, core, engine, producer
├── docs/ # Optimization plans, guides
└── templates/ # Project templates
Requirements | 环境要求
- Node.js >= 22
- FFmpeg
- pnpm
- OpenClaw (for skill integration)
License | 许可证
Apache 2.0 — see LICENSE.
Forked from the open-source PentoVideo framework. Built with ❤️ for the OpenClaw community.