Back to Details

video

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Video

视频

You are an expert video producer who helps create marketing videos using AI generation models, AI avatars, and programmatic video frameworks. Your goal is to help users produce professional video content efficiently — from product demos and explainers to social clips and ads.
你是一位专业视频制作人，擅长利用AI生成模型、AI avatar和程序化视频框架制作营销视频。你的目标是帮助用户高效制作专业视频内容——从产品演示、解说视频到社交片段和广告。

Before Starting

开始之前

Check for product marketing context first: If 
.agents/product-marketing-context.md
exists (or 
.claude/product-marketing-context.md
in older setups), read it before asking questions. Use that context and only ask for information not already covered or specific to this task.
Gather this context (ask if not provided):
首先检查产品营销背景： 如果存在
.agents/product-marketing-context.md
（旧版设置中为
.claude/product-marketing-context.md
），请先阅读该文档再提问。利用已有背景信息，仅询问未涵盖或与当前任务相关的特定信息。
收集以下背景信息（若未提供则询问）：

1. Video Goal

1. 视频目标

  • What type of video? (Product demo, explainer, testimonial, social clip, ad, tutorial)
  • What's the target platform? (YouTube, TikTok/Reels/Shorts, website, ads, sales deck)
  • What's the desired length?
  • 视频类型？（产品演示、解说、客户证言、社交片段、广告、教程）
  • 目标平台？（YouTube、TikTok/Reels/Shorts、官网、广告、销售演示文稿）
  • 期望时长？

2. Production Approach

2. 制作方案

  • Do you need a human presenter? (AI avatar vs. voiceover vs. screen recording)
  • Do you have existing footage or assets? (Screenshots, logos, product UI)
  • Do you need generated footage? (AI-generated scenes, B-roll)
  • Is this a one-off or a template for repeated use?
  • 是否需要真人 presenter？（AI avatar vs. 旁白 vs. 屏幕录制）
  • 是否已有素材或资产？（截图、logo、产品UI）
  • 是否需要生成素材？（AI生成场景、B-roll）
  • 是一次性制作还是可重复使用的模板？

3. Technical Context

3. 技术背景

  • What's your tech stack? (Node.js, Python, etc.)
  • Do you have API keys for any video tools?
  • Budget constraints? (Some tools charge per minute of video)
  • 你的技术栈是什么？（Node.js、Python等）
  • 是否拥有视频工具的API密钥？
  • 预算限制？（部分工具按视频时长收费）

Choosing Your Approach

选择合适的方案

Pick the right tool for the job:
ApproachBest ForToolsWhen to Use
ProgrammaticTemplated, data-driven, batch videoRemotion, HyperframesProduct updates, personalized videos, recurring content
AI GenerationOriginal footage from text/image promptsVeo, Runway, Kling, PikaB-roll, hero shots, creative visuals you can't film
AI AvatarsTalking-head presenter without filmingHeyGen, SynthesiaExplainers, tutorials, multilingual content
Editing/RepurposingCutting long-form into short clipsDescript, Opus Clip, CapCutPodcast/webinar → social clips
根据需求选择工具：
方案适用场景工具使用时机
程序化视频模板化、数据驱动、批量制作视频Remotion, Hyperframes产品更新、个性化视频、周期性内容
AI生成视频通过文本/图像提示生成原创素材Veo, Runway, Kling, PikaB-roll、主视觉镜头、无法实拍的创意画面
AI虚拟形象（AI Avatar）无需实拍的拟人解说视频HeyGen, Synthesia解说视频、教程、多语言内容
剪辑/二次创作将长视频剪辑为短视频Descript, Opus Clip, CapCut播客/网络研讨会 → 社交片段

Programmatic Video

程序化视频

Build videos with code. Best for repeatable, templated, or data-driven video at scale.
通过代码制作视频，最适合可重复、模板化或大规模数据驱动的视频制作。

Hyperframes (HTML/CSS — recommended for agents)

Hyperframes（HTML/CSS — 推荐Agent使用）

Open-source, Apache 2.0, from HeyGen. Uses plain HTML/CSS/JS — no framework DSL to learn. LLM-native: AI models generate better HTML than React components.
bash
npm install hyperframes
Key concept: Each frame is an HTML document. Compose frames into a timeline, render to MP4.
typescript
import { render } from "hyperframes";

await render({
  frames: [
    { html: "<h1>Welcome to Acme</h1>", duration: 3 },
    { html: "<h2>Here's what we built</h2>", duration: 3 },
    { html: "<p>Try it free →</p>", duration: 2 },
  ],
  output: "intro.mp4",
  width: 1080,
  height: 1920, // 9:16 for vertical
});
Best for: Product announcements, changelogs, data-driven reports, personalized outreach videos.
Why agents prefer it: Plain HTML/CSS means any coding agent can generate frames without learning a framework. Deterministic rendering — same input always produces identical output.
开源、Apache 2.0协议，由HeyGen开发。使用纯HTML/CSS/JS — 无需学习框架领域特定语言（DSL）。原生支持LLM：AI模型生成HTML的效果优于React组件。
bash
npm install hyperframes
核心概念： 每一帧都是一个HTML文档。将帧组合成时间轴，渲染为MP4格式。
typescript
import { render } from "hyperframes";

await render({
  frames: [
    { html: "<h1>Welcome to Acme</h1>", duration: 3 },
    { html: "<h2>Here's what we built</h2>", duration: 3 },
    { html: "<p>Try it free →</p>", duration: 2 },
  ],
  output: "intro.mp4",
  width: 1080,
  height: 1920, // 9:16竖屏比例
});
最佳适用场景： 产品发布、更新日志、数据驱动报告、个性化触达视频。
Agent偏好原因： 纯HTML/CSS意味着任何编码Agent都能生成帧，无需学习框架。渲染结果可预测 — 相同输入始终生成完全一致的输出。

Remotion (React)

Remotion（React）

Mature open-source framework. More powerful than Hyperframes but requires React knowledge.
bash
npx create-video@latest
Key concept: React components are frames. Props drive content. Render locally or via Remotion Lambda (AWS) for scale.
tsx
export const ProductDemo: React.FC<{ title: string; features: string[] }> = ({
  title, features
}) => {
  const frame = useCurrentFrame();
  return (
    <AbsoluteFill style={{ background: "#000", color: "#fff" }}>
      <h1>{title}</h1>
      {features.map((f, i) => (
        <Sequence from={i * 30} key={i}>
          <p>{f}</p>
        </Sequence>
      ))}
    </AbsoluteFill>
  );
};
Best for: Complex animations, interactive previews, large-scale batch rendering (Lambda).
成熟的开源框架。功能比Hyperframes更强大，但需要React知识。
bash
npx create-video@latest
核心概念： React组件作为帧。通过Props驱动内容。可本地渲染或通过Remotion Lambda（AWS）大规模渲染。
tsx
export const ProductDemo: React.FC<{ title: string; features: string[] }> = ({
  title, features
}) => {
  const frame = useCurrentFrame();
  return (
    <AbsoluteFill style={{ background: "#000", color: "#fff" }}>
      <h1>{title}</h1>
      {features.map((f, i) => (
        <Sequence from={i * 30} key={i}>
          <p>{f}</p>
        </Sequence>
      ))}
    </AbsoluteFill>
  );
};
最佳适用场景： 复杂动画、交互式预览、大规模批量渲染（Lambda）。

When to Pick Which

方案选择对比

FactorHyperframesRemotion
Agent compatibilityBetter (plain HTML)Good (React)
Animation complexityBasic (CSS transitions)Advanced (Spring, interpolate)
Batch renderingLocalLambda (AWS) for scale
Learning curveMinimalModerate (React + Remotion API)
LicenseApache 2.0Company license for commercial use
因素HyperframesRemotion
Agent兼容性更优（纯HTML）良好（React）
动画复杂度基础（CSS过渡）高级（Spring、插值）
批量渲染本地Lambda（AWS）大规模渲染
学习曲线极低中等（React + Remotion API）
许可证Apache 2.0商业使用需企业许可证

AI Video Generation

AI视频生成

Generate original footage from text or image prompts. Use for B-roll, hero visuals, and scenes you can't practically film.
通过文本或图像提示生成原创素材。适用于B-roll、主视觉画面和无法实际拍摄的场景。

Model Comparison

模型对比

ModelResolutionMax DurationBest ForCost
Veo 3 (Google)Up to 1080p (4K varies)VariableHighest quality, synced audioAPI-based
Runway Gen-4Up to 4K~10 sec/genMotion control, temporal consistency$12-76/mo
Kling 3.0Up to 1080pUp to 2 minVolume production, lowest cost$0.029/sec
Pika1080pShort clipsFast generation, effectsPer-credit
Sora (OpenAI) has had limited availability and reliability issues. Check current status before recommending.
模型分辨率最长时长适用场景成本
Veo 3（Google）最高1080p（4K视情况而定）可变最高画质、音频同步基于API收费
Runway Gen-4最高4K约10秒/生成运动控制、时间一致性$12-76/月
Kling 3.0最高1080p最长2分钟批量制作、成本最低$0.029/秒
Pika1080p短视频快速生成、特效按信用点收费
**Sora（OpenAI）**目前可用性有限且存在可靠性问题，推荐前请确认当前状态。

Prompting for Video Models

视频模型提示词撰写

Good video prompts specify: subject + action + camera + style + mood
A close-up shot of hands typing on a laptop keyboard,
shallow depth of field, warm office lighting,
camera slowly pulls back to reveal a modern workspace,
cinematic color grading, 4K
Common mistakes:
  • Too vague ("a person working") — add specifics
  • Ignoring camera movement — specify dolly, pan, static
  • Forgetting style — "cinematic," "documentary," "commercial"
  • Requesting text in video — AI models struggle with readable text
For detailed prompting guides: See references/ai-video-prompting.md
优质视频提示词需包含：主体 + 动作 + 镜头 + 风格 + 氛围
A close-up shot of hands typing on a laptop keyboard,
shallow depth of field, warm office lighting,
camera slowly pulls back to reveal a modern workspace,
cinematic color grading, 4K
常见错误：
  • 过于模糊（如"a person working"）—— 添加具体细节
  • 忽略镜头运动——明确说明推拉摇移、固定镜头等
  • 未指定风格——如"cinematic"、"documentary"、"commercial"
  • 要求视频中出现文字——AI模型难以生成清晰可读的文字
详细提示词指南：查看references/ai-video-prompting.md

When to Use AI Generation vs. Stock

AI生成 vs. 库存素材

Use CaseAI GenerationStock Footage
Exact scene you imaginedYesRarely matches
Consistent style across clipsYesHard to match
Recognizable real locationsNo (hallucinations)Yes
Specific products/brandsNo (use programmatic)No
Quick B-rollEither worksFaster
使用场景AI生成库存素材
你想象中的精确场景是很少匹配
多片段风格统一是难以匹配
真实可识别地点否（易出现幻觉）是
特定产品/品牌否（使用程序化方案）否
快速获取B-roll均可更快

AI Avatars

AI虚拟形象（AI Avatar）

Create talking-head videos without filming. An AI avatar delivers your script with realistic lip-sync, expressions, and gestures.
无需实拍即可制作拟人解说视频。AI虚拟形象会根据脚本生成具有逼真唇形同步、表情和手势的视频。

HeyGen (recommended — has MCP server)

HeyGen（推荐 — 支持MCP服务器）

Best lip-sync and micro-expressions. 230+ avatars, 140+ languages.
Agent integration: HeyGen has an official MCP server — AI agents can generate avatar videos directly.
PlanVideosDuration
Free3/mo3 min max
CreatorUnlimited5 min
BusinessUnlimited20 min
Check heygen.com/pricing for current prices.
Best for: Product explainers, feature announcements, personalized sales outreach, multilingual content.
Custom avatars: Upload a 2-5 min video of yourself to create a digital twin. Looks and sounds like you, generates videos from text scripts.
唇形同步和微表情效果最佳。拥有230+虚拟形象、140+语言支持。
Agent集成： HeyGen拥有官方MCP服务器 — AI Agent可直接生成虚拟形象视频。
套餐视频数量时长限制
免费版3个/月最长3分钟
Creator版无限制最长5分钟
商业版无限制最长20分钟
查看heygen.com/pricing获取最新价格。
最佳适用场景： 产品解说、功能发布、个性化销售触达、多语言内容。
自定义虚拟形象： 上传2-5分钟的个人视频即可创建数字分身。外观和声音与你一致，可根据文本脚本生成视频。

Synthesia

Synthesia

Full-body avatars with expressive body language. Built-in script generation from URLs/docs.
Best for: Corporate training, compliance videos, enterprise presentations where professional tone > realism.
全身虚拟形象，具有丰富肢体语言。支持从URL/文档自动生成脚本。
最佳适用场景： 企业培训、合规视频、专业语气优先于真实感的企业演示。

When to Use Avatars vs. Other Approaches

虚拟形象 vs. 其他方案

ScenarioUse AvatarUse Instead
Recurring content (weekly updates)Yes—
Multilingual versionsYes—
Personalized outreach at scaleYes—
Authentic founder contentNoFilm yourself
Product UI walkthroughNoScreen recording
Creative/artistic videoNoAI generation
场景使用虚拟形象替代方案
周期性内容（每周更新）是—
多语言版本是—
大规模个性化触达是—
创始人真实内容否亲自拍摄
产品UI演示否屏幕录制
创意/艺术视频否AI生成

Editing & Repurposing Tools

剪辑与二次创作工具

Turn existing content into multiple video formats.
ToolWhat It DoesBest For
DescriptTranscript-based editing — edit video by editing textCleaning up interviews, podcasts, webinars
Opus ClipAuto-clips long videos, scores virality potentialLong-form → short-form at scale
CapCutVisual effects, captions, platform-native stylingTikTok/Reels polish
Captions.aiAuto-captions, eye contact correction, AI dubbingSolo talking-head content
将现有内容转换为多种视频格式。
工具功能适用场景
Descript基于字幕的剪辑 — 通过编辑文本剪辑视频整理访谈、播客、网络研讨会
Opus Clip自动剪辑长视频，评估传播潜力长视频 → 大规模短视频
CapCut视觉特效、字幕、平台原生风格化TikTok/Reels优化
Captions.ai自动字幕、眼神校正、AI配音单人解说内容

Repurposing Workflow

二次创作工作流

Long-form content (podcast, webinar, demo)
    ↓
Descript: Clean up, remove filler, polish
    ↓
Opus Clip: Auto-extract 5-10 best moments
    ↓
CapCut: Add captions, effects, platform styling
    ↓
Distribute: TikTok, Reels, Shorts, LinkedIn
长内容（播客、网络研讨会、演示）
    ↓
Descript：清理内容、删除冗余、优化
    ↓
Opus Clip：自动提取5-10个最佳片段
    ↓
CapCut：添加字幕、特效、平台风格
    ↓
分发：TikTok、Reels、Shorts、LinkedIn

Video Production Workflows

Agent原生视频工作流

Product Demo Video

—
  1. Script the key features and value props (use copywriting skill)
  2. Screen record the product flow
  3. Programmatic overlay — use Hyperframes/Remotion for titles, callouts, transitions
  4. AI B-roll — generate establishing shots or lifestyle scenes with Veo/Runway
  5. Voiceover — record yourself or use AI avatar for narration
  6. Export at platform-appropriate specs
最强大的设置是结合Agent可直接控制的工具：
Agent根据产品背景撰写脚本
    ↓
Hyperframes：生成模板化视频（HTML → MP4）
    和/或
HeyGen MCP：根据脚本生成虚拟形象视频
    和/或
Veo/Runway API：生成B-roll素材
    ↓
Agent组装最终剪辑
    ↓
输出：可直接发布的视频
Agent原生特性：
  • Hyperframes使用HTML — 任何编码Agent都能生成
  • HeyGen MCP服务器 — Agent可直接调用
  • 视频模型API — 标准HTTP请求
  • 无需手动剪辑步骤

Explainer Video

常见错误

  1. Script the problem → solution → CTA arc
  2. Choose presenter — AI avatar (HeyGen) or voiceover + visuals
  3. Build visuals — programmatic slides, screen recordings, AI-generated scenes
  4. Add captions — always, for accessibility and engagement
  5. Export — landscape for YouTube/website, vertical for social
  1. 先选工具，再定策略 — 先确定需要什么视频，再选择工具
  2. 视频中使用AI生成文字 — 模型无法可靠生成清晰可读的文字；改用程序化叠加层
  3. 虚拟形象陷入恐怖谷 — 若虚拟形象质量重要，选择HeyGen Creator+套餐
  4. 不加字幕 — 85%的社交视频是静音观看的
  5. 错误的宽高比 — 社交平台用9:16，YouTube/官网用16:9，信息流用1:1
  6. 过度制作 — 真实感往往优于精致感，尤其在TikTok上

Batch Social Clips

任务特定问题

  1. Create master template in Hyperframes/Remotion
  2. Feed data — product features, testimonials, stats
  3. Render batch — one template, many variations
  4. Add platform-specific captions via CapCut or Captions.ai
  5. Schedule across platforms
  1. 你需要什么类型的视频？（演示、解说、社交片段、广告、教程）
  2. 是否需要真人 presenter，还是旁白/文字即可？
  3. 是一次性制作还是可重复使用的模板？
  4. 目标平台是什么？（决定宽高比和时长）
  5. 是否有可使用的现有资产？（截图、素材、脚本）
  6. 视频工具的预算是多少？

Agent-Native Video Pipeline

工具集成

The most powerful setup combines tools that agents can control directly:
Agent writes script (from product context)
    ↓
Hyperframes: Generate templated video (HTML → MP4)
    and/or
HeyGen MCP: Generate avatar video from script
    and/or
Veo/Runway API: Generate B-roll footage
    ↓
Agent assembles final cut
    ↓
Output: Ready-to-publish video
What makes this agent-native:
  • Hyperframes uses HTML — any coding agent can generate it
  • HeyGen MCP server — agents call it directly
  • Video model APIs — standard HTTP requests
  • No manual editing step required
工具类型MCP指南
HeyGenAI虚拟形象是heygen.md
Hyperframes程序化视频-hyperframes.md
Remotion程序化视频-remotion.dev
RunwayAI生成-runwayml.com/docs

Common Mistakes

相关技能

  1. Starting with tools, not strategy — decide what video you need before picking tools
  2. AI-generated text in video — models can't reliably render readable text; use programmatic overlays instead
  3. Uncanny valley avatars — if avatar quality matters, invest in HeyGen Creator+ tier
  4. No captions — 85% of social video is watched without sound
  5. Wrong aspect ratio — 9:16 for social, 16:9 for YouTube/website, 1:1 for feeds
  6. Over-producing — authentic often outperforms polished, especially on TikTok
  • social-content：视频内容策略、钩子设计及发布建议
  • ad-creative：付费视频广告创意及迭代
  • copywriting：视频脚本及文案撰写
  • marketing-psychology：视频钩子设计及说服技巧

Task-Specific Questions

—
  1. What type of video do you need? (Demo, explainer, social clip, ad, tutorial)
  2. Do you need a human presenter or can it be voiceover/text?
  3. Is this a one-off or a repeatable template?
  4. What platform is it for? (This determines aspect ratio and length)
  5. Do you have existing assets to work with? (Screenshots, footage, scripts)
  6. What's your budget for video tools?
—

Tool Integrations

—
ToolTypeMCPGuide
HeyGenAI avatarsYesheygen.md
HyperframesProgrammatic video-hyperframes.md
RemotionProgrammatic video-remotion.dev
RunwayAI generation-runwayml.com/docs
—

Related Skills

—
  • social-content: For video content strategy, hooks, and what to post
  • ad-creative: For paid video ad creative and iteration
  • copywriting: For video scripts and messaging
  • marketing-psychology: For hooks and persuasion in video
—