Token导航 LogoToken导航

Agent Skill · 研究检索

ai-video-skill(AI 视频技能)

用于辅助视频生成、动画合成、脚本化剪辑或 Remotion 等视频项目开发。它适合让 Agent 组织镜头、生成素材说明、维护合成代码或排查渲染问题。使用时需要确认分辨率、时长、素材路径和导出格式;涉及外部素材、人物肖像或商业发布时,应先核对版权授权和内容审核要求。

ai-video-skill 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为接触敏感数据;许可证为 MIT-0,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 23,296。

总安装 4,304周安装 183GitHub Stars 1下载 1,508核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.0

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:ai-video-skill(AI 视频技能)
来源仓库:https://github.com/dhivalogu/ai-video-skill
安装命令:
openclaw skills install ai-video-skill
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install ai-video-skill

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于 AI 视频创作,生成带旁白、视觉效果与字幕的多平台适配视频。支持 YouTube Shorts、TikTok 等短格式与长视频内容制作。使用时需确认素材版权与平台规范,涉及商业用途应审核合规性。

安装方式:通过 clawhub 安装,命令为 openclaw skills install ai-video-skill。注意:涉及人物或品牌内容时,需额外核对授权与真实性。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install ai-video-skill;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

ai-video-skill 怎么安装?

在终端执行 openclaw skills install ai-video-skill。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

ai-video-skill 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

ai-video-skill 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

ai-video-skill 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

ai-video-skill 用的人多吗?

公开数据是累计安装 4,304 次,近一周安装 183 次,下载 1,508 次,GitHub 1 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

ai-video-skill 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
ai-video-skill

AI 视频技能 · 本页

4,3041接触敏感数据MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
aituber-api
description
>
license
MIT
compatibility
allowed-tools
Bash(curl:*) WebFetch Read Write
metadata
openclaw
emoji
\F3AC
homepage
https://aituber.app
primaryEnv
AITUBER_API_KEY
requires
env
bins

AITuber API

Create AI videos from a script or idea. From 15-second Shorts to 7-minute long-form content. The API handles voice narration (1,300+ voices, any language), AI-generated visuals, word-synced captions, MP4 export, and publishing to YouTube, TikTok, and Instagram.

Base URL: https://app.aituber.app/api/v1 OpenAPI spec: https://app.aituber.app/api/v1/openapi.json API docs: https://app.aituber.app/api-docs

Tip: For the most up-to-date parameter details, examples, and response schemas, refer to the interactive API docs at https://app.aituber.app/api-docs when possible. The reference below covers everything you need to get started, but the docs always reflect the latest API.

Authentication

All endpoints except GET /voices require a Bearer token. Create an API key in the AITuber dashboard at https://app.aituber.app/dashboard/api-keys.

Authorization: Bearer <AITUBER_API_KEY>

Store the key in the AITUBER_API_KEY environment variable.

Complete Workflow

The typical flow is: pick voice, generate video, poll status, export, download.

Step 1: Pick a voice

curl "https://app.aituber.app/api/v1/voices?gender=male&accent=American"

No auth required. Returns an array of voice objects. Use the id field as voiceId when generating.

Step 2: Generate a video

From a script (you write the narration):

curl -X POST "https://app.aituber.app/api/v1/videos/generate" \
  -H "Authorization: Bearer $AITUBER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "script": "The human brain contains roughly 86 billion neurons. Every thought, memory, and emotion is the result of electrical signals racing through this incredible network.",
    "voiceId": "nPczCjzI2devNBz1zQrb",
    "imageStyleId": "cinematic",
    "aspectRatio": "9:16"
  }'

From an idea (AI writes the script):

curl -X POST "https://app.aituber.app/api/v1/videos/generate" \
  -H "Authorization: Bearer $AITUBER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "script": "5 mind-blowing facts about black holes",
    "inputType": "idea",
    "expectedDurationSeconds": 60
  }'

Returns { "videoId": "uuid", "status": "pending" }.

Step 3: Poll for completion

curl "https://app.aituber.app/api/v1/videos/VIDEO_ID" \
  -H "Authorization: Bearer $AITUBER_API_KEY"

Poll every 5-10 seconds. Generation typically takes 1-3 minutes. Wait until status is completed (or failed).

Step 4: Export to MP4

curl -X POST "https://app.aituber.app/api/v1/exports" \
  -H "Authorization: Bearer $AITUBER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "videoId": "VIDEO_ID", "resolution": "1080p" }'

Exporting is free. Requires an active paid subscription. Poll GET /videos/VIDEO_ID and check exportStatus until it is completed (typically 30 seconds to 5 minutes).

Step 5: Download the MP4

curl "https://app.aituber.app/api/v1/exports/download?videoId=VIDEO_ID" \
  -H "Authorization: Bearer $AITUBER_API_KEY"

Returns { "url": "https://...", "videoId": "..." }. The URL is a signed temporary link that expires in 2 minutes. Download immediately.

Endpoints Reference

GET /voices

List available AI voices. No auth required.

ParamTypeDescription
genderstringFilter: male, female
accentstringFilter: American, British, Australian, Indian
useCasestringFilter: narration, conversational, characters, social_media
searchstringSearch by name or description

Response: array of { id, name, description, previewUrl, gender, age, accent, useCase, optimizedFor }.

POST /videos/generate

Create a new video. Returns { videoId, status }.

Video types (all use this same endpoint):

  1. Faceless narration (images) - Default. AI images with Ken Burns animation. Just send a script or idea.
  2. Faceless narration (video clips) - AI video clips. Set mediaType: "video".
  3. Stock footage - Real stock footage matched to narration. Set mediaType: "stock".
  4. Skeleton template - Viral "what happens if..." X-ray style. Set templateId: "skeleton". Template handles mediaType/style.
  5. Character template - Character-driven stories. Set templateId: "character", inputType: "idea". Only idea mode (AI writes the script for character consistency).

Examples:

// Idea to Video (faceless images, AI writes the script)
{ "script": "5 facts about black holes", "inputType": "idea", "expectedDurationSeconds": 60 }

// Script to Video (video clips, you write the script)
{ "script": "The ocean covers over 70 percent of Earth...", "mediaType": "video" }

// Skeleton template
{ "script": "What happens if you eat 100 bananas", "templateId": "skeleton" }

// Character template
{ "script": "A detective solves a mystery in Tokyo", "templateId": "character", "inputType": "idea", "expectedDurationSeconds": 90 }

// Script with visual control (brackets tell AI what to show)
{ "script": "[A dark forest at night] The wind howled through the trees. [Glowing eyes in the shadows] Something was watching.", "imageStyleId": "cinematic" }
ParamTypeRequiredDescription
scriptstringYesNarration text (script mode) or topic (idea mode). The AI generates visuals automatically. For more control, optionally add visual instructions in brackets: [A dark forest] The wind howled. Each [bracket] tells the AI what to show for that scene.
inputTypestringNoscript (default) or idea. Character template requires idea.
voiceIdstringNoVoice ID from GET /voices. Default: "Adam"
voiceSpeednumberNo0.7 to 1.2. Default: 1.0
mediaTypestringNoimages (default), video, or stock. Ignored when using templates.
aspectRatiostringNo9:16 (default), 16:9, or 1:1
imageQualitystringNoFor images mediaType: basic (1 credit/image), good (3), premium (9), max (45). Ignored when using templates.
imageStyleIdstringNoVisual style: photorealistic (default), cinematic, anime, 3d-pixar, watercolor, comic-book, etc. Ignored when using templates.
captionStyleIdstringNowrap-1 (default), hormozi, classic, karaoke, box, minimal, handwritten, neon. Create custom styles in the dashboard.
captionsEnabledbooleanNoDefault: true
captionPositionstringNobottom (default), center, top
expectedDurationSecondsnumberNoTarget length in seconds (15-420). Required for idea mode. Max 420 (7 minutes).
templateIdstringNoskeleton or character. When using a template, do NOT send mediaType, imageStyleId, or imageQuality. The template handles these automatically.
videoQualitystringNoFor video mediaType: basic, good (default), premium

GET /videos

List your videos. Returns array sorted newest first.

ParamTypeDescription
limitnumberMax results, 1-100. Default: 50

GET /videos/{id}

Get a single video by ID. Use for polling generation and export status.

GET /subscription

Check your plan and credit balance. Returns { plan, status, credits, monthlyCredits }.

POST /exports

Export a completed video to MP4. Free, but requires an active paid subscription.

ParamTypeRequiredDescription
videoIdstringYesVideo ID (must have status: completed)
resolutionstringNo1080p (default) or 4k

GET /exports/download

Get a temporary download URL for a rendered MP4.

ParamTypeDescription
videoIdstringFinds the latest completed export for this video

The returned URL expires in 2 minutes. Download immediately.

Credit Costs

OperationCost
Image (basic)1 credit/image
Image (good)3 credits/image
Image (premium)9 credits/image
Image (max)45 credits/image
Audio narration~50 credits/minute
Video clip (4s)75 credits
Video clip (8s)150 credits
Export to MP4Free

A typical 60-second video with basic image quality costs approximately 75 credits.

Error Handling

HTTP StatusMeaning
401Missing or invalid API key
402Not enough credits. Response includes creditsRequired and creditsAvailable
403Feature requires an active paid subscription
404Resource not found
400Bad request (check error message for details)

Full Example (bash)

# 1. Check credits
curl -s "https://app.aituber.app/api/v1/subscription" \
  -H "Authorization: Bearer $AITUBER_API_KEY" | jq .

# 2. Browse voices
curl -s "https://app.aituber.app/api/v1/voices?gender=female&useCase=narration" | jq '.[0:3]'

# 3. Generate video
VIDEO_ID=$(curl -s -X POST "https://app.aituber.app/api/v1/videos/generate" \
  -H "Authorization: Bearer $AITUBER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "script": "Did you know that octopuses have three hearts and blue blood?",
    "voiceId": "EXAVITQu4vr4xnSDxMaL",
    "imageStyleId": "cinematic",
    "aspectRatio": "9:16"
  }' | jq -r '.videoId')

echo "Video ID: $VIDEO_ID"

# 4. Poll until completed
while true; do
  STATUS=$(curl -s "https://app.aituber.app/api/v1/videos/$VIDEO_ID" \
    -H "Authorization: Bearer $AITUBER_API_KEY" | jq -r '.status')
  echo "Status: $STATUS"
  if [ "$STATUS" = "completed" ] || [ "$STATUS" = "failed" ]; then break; fi
  sleep 10
done

# 5. Export to MP4
curl -s -X POST "https://app.aituber.app/api/v1/exports" \
  -H "Authorization: Bearer $AITUBER_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{\"videoId\": \"$VIDEO_ID\"}"

# 6. Poll export status
while true; do
  EXPORT_STATUS=$(curl -s "https://app.aituber.app/api/v1/videos/$VIDEO_ID" \
    -H "Authorization: Bearer $AITUBER_API_KEY" | jq -r '.exportStatus')
  echo "Export: $EXPORT_STATUS"
  if [ "$EXPORT_STATUS" = "completed" ] || [ "$EXPORT_STATUS" = "failed" ]; then break; fi
  sleep 10
done

# 7. Download
DOWNLOAD_URL=$(curl -s "https://app.aituber.app/api/v1/exports/download?videoId=$VIDEO_ID" \
  -H "Authorization: Bearer $AITUBER_API_KEY" | jq -r '.url')

curl -L -o video.mp4 "$DOWNLOAD_URL"
echo "Downloaded: video.mp4"

平台分布

OpenClaw

73.07%
按下载量换算1,102

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills