Token导航 LogoToken导航

Agent Skill · 研究检索

video-production(视频制作)

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。

video-production 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为接触敏感数据;许可证信息还不明确,第三方扫描有未通过项,按安装量在研究检索类的 73,675 个 Skill 里排第 23,369。

总安装 4,284周安装 184GitHub Stars 11下载 1,501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
unknown

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:video-production(视频制作)
来源仓库:https://github.com/casper-studios/casper-marketplace
仓库路径:skills/video-production
安装命令:
npx skills add https://github.com/casper-studios/casper-marketplace --skill video-production
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 npx skills 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。

skills.shnpx skills
npx skills add https://github.com/casper-studios/casper-marketplace --skill video-production

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。

video-production 组装课程视频片段并生成 YouTube 发布所需的标题页与时间轴描述。支持从 Google Drive 批量下载素材、拼接成片与自动生成字幕文件。适用于在线教育、培训课件与产品演示等多媒体内容制作流水线。需配置视频分辨率、时长与版权素材路径,商业用途需额外审核授权文件。

适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/video-production

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 Codex、Claude、Cursor、Gemini CLI;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 npx skills add https://github.com/casper-studios/casper-marketplace --skill vide…;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

video-production 怎么安装?

在终端执行 npx skills add https://github.com/casper-studios/casper-marketplace --skill video-producti…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

video-production 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。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。

video-production 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。

video-production 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

video-production 用的人多吗?

公开数据是累计安装 4,284 次,近一周安装 184 次,下载 1,501 次,GitHub 11 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。

video-production 安全吗?

第三方扫描有未通过项,风险较高。除非你确认过源码,否则不建议直接安装到常用环境。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

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。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 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 Codex,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。

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。
  • 需要明确授权才能用到公司项目里的场景。
  • 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
video-production

视频制作 · 本页

4,28411接触敏感数据未明确未通过

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

Video Production

Overview

Assemble course videos from individual lesson files with title slides and auto-generated timestamps for YouTube.

Quick Decision Tree

What do you need?
│
├── Full course assembly (end-to-end)
│   └── references/workflow.md
│   └── Combines all scripts below
│
├── Download videos from Drive
│   └── Script: scripts/gdrive_video_download.py
│
├── Create title slides
│   └── Script: scripts/create_title_slides.py
│
├── Stitch videos together
│   └── Script: scripts/stitch_videos.py
│
└── Generate YouTube description
    └── Script: scripts/generate_youtube_description.py

Environment Setup

Google Drive OAuth (same as google-workspace skill).

System Requirements

  • FFmpeg installed and in PATH
  • Python 3.9+

Complete Workflow

# Full course assembly from Drive folder
python scripts/stitch_videos.py \
  --folder "https://drive.google.com/drive/folders/xxx" \
  --output "Complete Course.mp4" \
  --slide-duration 3

Pipeline Steps

  1. Download - Get all videos from Drive folder
  2. Parse Titles - Extract clean names from [e1] Intro format
  3. Get Metadata - FFprobe for duration/resolution
  4. Generate Slides - Title card for each lesson
  5. Build Concat List - video1 → slide2 → video2 →...
  6. Stitch with FFmpeg - Concatenate all segments
  7. Calculate Timestamps - Track cumulative duration
  8. Generate Description - YouTube-ready markdown

Outputs

FileDescription
{output_name}.mp4Final stitched video
youtube_description.mdTimestamped description
metadata.jsonProcessing info

Performance

InputTimeOutput Size
5 videos (30 min)~5 min~1.5 GB
10 videos (1 hr)~10 min~3 GB
20 videos (2 hr)~20 min~6 GB

Security Notes

Credential Handling

  • Google OAuth credentials for Drive access (see google-workspace skill)
  • mycreds.txt and client_secrets.json never committed to git
  • No additional API keys required for local video processing

Data Privacy

  • All video processing happens locally using FFmpeg
  • No video content is uploaded to external cloud services
  • Source videos downloaded from Google Drive to local .tmp/
  • Final videos stored locally until manually uploaded
  • Metadata JSON contains file names and timestamps only

Access Scopes

  • Google Drive: drive.readonly sufficient for downloading
  • Google Drive: drive required for uploading final videos
  • No external video processing APIs used

Compliance Considerations

  • Local Processing: All encoding/stitching done locally (privacy-preserving)
  • No Cloud Upload: Videos never leave your machine during processing
  • Content Rights: Ensure you have rights to source video content
  • Course Content: Verify licensing for educational content distribution
  • YouTube ToS: Generated descriptions comply with YouTube guidelines
  • Storage: Large video files require adequate local disk space
  • Cleanup: Remove temporary files from .tmp/ after processing

Troubleshooting

Common Issues

Issue: FFmpeg not found

Symptoms: "FFmpeg not found" or "command not found: ffmpeg" Cause: FFmpeg not installed or not in system PATH Solution:

  • Install FFmpeg: brew install ffmpeg (macOS) or download from ffmpeg.org
  • Verify installation: ffmpeg -version
  • Add FFmpeg to PATH if installed in non-standard location
  • Restart terminal after installation
Issue: Codec mismatch / incompatible videos

Symptoms: "Non-monotonous DTS" or codec errors during stitching Cause: Source videos have different codecs, resolutions, or frame rates Solution:

  • Re-encode all source videos to the same format before stitching
  • Use FFmpeg to normalize: ffmpeg -i input.mp4 -c:v libx264 -c:a aac output.mp4
  • Ensure consistent resolution (e.g., all 1920x1080)
  • Match frame rates across all videos (e.g., all 30fps)
Issue: Audio out of sync

Symptoms: Audio drifts from video over time Cause: Inconsistent frame rates or variable frame rate sources Solution:

  • Use constant frame rate for all source videos
  • Re-encode with -vsync cfr flag
  • Avoid mixing video from different sources/devices
  • Check audio sample rates match across files
Issue: Insufficient disk space

Symptoms: "No space left on device" or incomplete output Cause: Not enough free space for video processing Solution:

  • Check available disk space: df -h
  • Clear .tmp/ directory of old files
  • Move large source videos to external drive
  • Process fewer videos at once
Issue: Google Drive download fails

Symptoms: Videos fail to download from Drive folder Cause: OAuth issue, permissions, or network timeout Solution:

  • Verify Google OAuth credentials (see google-workspace skill)
  • Check folder sharing permissions
  • Try downloading single file first to test
  • Check for network connectivity issues
Issue: Title slides not generating

Symptoms: Missing title cards in final video Cause: Font or image generation issue Solution:

  • Verify ImageMagick or Pillow is installed
  • Check font files exist if custom fonts specified
  • Review title text for special characters
  • Try with default font settings first

Resources

  • references/workflow.md - Complete video course workflow

Integration Patterns

Full Course Pipeline

Skills: google-workspace → video-production → google-workspace Use case: End-to-end course video assembly Flow:

  1. Download lesson videos from Google Drive folder
  2. Generate title slides and stitch all videos together
  3. Upload final video back to Drive and generate YouTube description

Transcript to Timestamps

Skills: transcript-search → video-production Use case: Generate YouTube descriptions from meeting recordings Flow:

  1. Search transcripts for relevant meetings
  2. Extract topic timestamps from transcript
  3. Generate formatted YouTube description with chapter markers

Content to Title Slides

Skills: content-generation → video-production Use case: Create branded title cards for videos Flow:

  1. Generate title slide images with content-generation
  2. Export slides in video-compatible format
  3. Insert title slides between video segments

平台分布

Codex

37.04%
按下载量换算556

Claude

29.14%
按下载量换算437

Cursor

17.87%
按下载量换算268

Gemini CLI

9.69%
按下载量换算145

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

未通过

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。来源安全扫描存在提醒或未通过项,安装前请自行复核。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