AITutor – 类似于 vimtutor，但用于 AI 辅助编码。
我创建了一个互动终端教程，教授人工智能编程概念——上下文窗口、工具、MCP、钩子、子代理等。<p>我每天使用Claude Code、Cursor和GitHub Copilot，并且我注意到同样的现象：大多数开发者（包括我自己，起初也是）在使用这些工具时，并没有真正理解它们的底层工作原理。<br>什么是上下文窗口？<br>为什么人工智能会“忘记”某些东西？<br>MCP实际上是做什么的？<br>为什么有些提示比其他提示效果更好？<p>AITutor就像这个领域的vimtutor。它包含15节课程，涵盖初级/中级/高级三个层次，每节课都有理论知识、可以互动的ASCII可视化以及测验。它完全在你的终端中运行。<p>使用Go和Charm生态系统（Bubbletea、Lipgloss）构建。<p>试试吧：npx @aitutor/cli@latest<br>或者：brew tap naorpeled/tap && brew install aitutor<p>GitHub链接：https://github.com/naorpeled/aitutor<p>欢迎对课程内容提供反馈——缺少什么，哪里有问题，哪些概念是你希望有人早些向你解释的。
查看原文
I built an interactive terminal tutorial that teaches AI coding concepts — context windows, tools, MCP, hooks, subagents, and more.<p>I use Claude Code, Cursor, and GitHub Copilot daily, and I kept noticing the same thing: most developers (myself included, at first) use these tools without really understanding how they work under the hood.
What's a context window?
Why did the AI "forget" something?
What does MCP actually do?
Why do some prompts work better than others?<p>AITutor is like vimtutor for this stuff. 15 lessons across beginner/intermediate/advanced tiers, each with theory, an interactive ASCII visualization you can play with, and a quiz. It runs entirely in your terminal.<p>Built with Go and the Charm ecosystem (Bubbletea, Lipgloss).<p>Try it: npx @aitutor/cli@latest
Or: brew tap naorpeled/tap && brew install aitutor<p>GitHub: https://github.com/naorpeled/aitutor<p>Would love feedback on the curriculum — what's missing, what's wrong, what concepts you wish someone had explained to you earlier.