将视频课程或播放列表转换为交互式、基于证据的技能,供 Claude Code 和 Codex 用于教学、测验、参考和实际应用。

Stars

11

7 天增长

暂无数据

Fork 数

0

开放 Issue

0

开源协议

MIT

最近更新

2026-07-30

AI 仓库情报摘要
FR-AI / ANALYSIS

为什么值得关注

将被动观看视频转变为 AI 编码助手中的主动自适应学习体验,利用多模态理解并自动创建课程专用技能,无需手动配置。

适合谁使用

  • 使用 Claude Code 或 Codex 的开发者
  • 希望从视频课程中提取可操作知识的自学者
  • 创建交互式课程材料的教师
  • 将视频培训融入编码工作流程的团队

典型使用场景

  • 通过检查和个性化教学互动学习课程
  • 练习题目并获得答案点评
  • 将课程方法论应用于实际仓库或项目
  • 将课程用作特定故障模式或建议的参考

项目优势

  • 基于证据,结合语音、字幕和视觉状态的时间戳溯源
  • 多模态理解,对齐语音、章节、场景变化、OCR 和画面
  • 隔离的私有运行时,不干扰用户项目,并可恢复中断的工作
  • 尊重隐私和版权:原始媒体保留在本地,不绕过 DRM,身份验证可选用

使用前须知

  • 需要 Python 3.11–3.13、FFmpeg 及 ffprobe,首次设置需要网络
  • 依赖于专有的 Claude Code 或 Codex 宿主,不是独立工具
  • 媒体处理限制为 720p 分析视频,并有明确的安全上限,不适合存档级复制

README 快速开始

video-to-skill

Turn a video, playlist, or course into an evidence-grounded Skill that Claude Code or Codex can teach from, quiz you on, use as a reference, and apply to real work.

Install

Paste this request into Claude Code or Codex:

Install video-to-skill globally as a standalone Skill from https://github.com/Lum1104/video-to-skill. Prepare and verify its private runtime, then confirm that video-to-skill is available.

Or install it for both hosts with the open Agent Skills installer:

npx skills add Lum1104/video-to-skill --skill video-to-skill --global --agent claude-code --agent codex --yes

The command-line option requires Node.js and npx. Start a new host session after installation.

On first use, the Skill creates a private Python runtime and downloads its third-party dependencies from PyPI; that setup needs Python 3.11–3.13 and network access. Processing media also requires FFmpeg with ffprobe. If a course later needs local speech recognition or OCR, the host enables that capability inside the same private runtime; you do not create a virtual environment or run a package manager yourself.

Create a course Skill

Give the generator a video, playlist, course URL, or local media file:

HostExample
Claude Code/video-to-skill https://www.youtube.com/playlist?list=...
Codex$video-to-skill https://www.youtube.com/playlist?list=...

You can add the desired outcome naturally:

Build a Skill that teaches this course and can apply its method to my projects.

One invocation inventories the accessible course, acquires and investigates the evidence, builds and validates a course-specific Skill, and installs it for the active host. Interrupted work resumes from its private evidence workspace.

Use the generated Skill

If the generated Skill is named react-performance-course, use /react-performance-course in Claude Code or replace the leading / with $ in Codex:

GoalExample request
Learn/react-performance-course Check what I know, then teach me the first topic.
Practice/react-performance-course Give me an exercise and review my attempt.
Apply/react-performance-course Inspect this repository using the course's method.
Reference`/react-performance-course What did the course rec

项目描述

Turn videos and courses into evidence-grounded Agent Skills for Claude Code and Codex.

相关仓库与替代方案

根据分类、Topic 和编程语言匹配的相似项目。

gnipbao
精选
gnipbao GitHub avatar

story-to-handdrawn-video

A Remotion-based tool that converts Chinese story text or ordered hand-drawn images into vertical hand-drawn diary-comic animation with handwritten captions, left-to-right reveals, optional page-curl transitions, and silent H.264 output for post-production dubbing.

AI 与机器学习AI 智能体
683
MoonshotAI
精选
MoonshotAI GitHub avatar

Kimi-K3

Kimi K3 is an open-weight, 2.8T-parameter native multimodal agentic model with a 1M-token context window, designed for frontier coding, knowledge work, and reasoning tasks.

AI 与机器学习AI 智能体
3,348
yetone
精选
yetone GitHub avatar

kill-ai-slop

Kill AI Slop is a multilingual field guide and agent skill that catalogues 33 common AI-generated design tells with interactive before-after demos, and provides a tool to scan and fix web projects.

AI 与机器学习AI 智能体
805