在 Codex CLI 里为学习/知识梳理类问答自动生成视觉化笔记 — 包括流程图、架构图、时序图、对比表、状态机、学习路径、交互卡片等。默认产出 Obsidian 兼容的 markdown 笔记(内嵌 SVG 或 dataviewjs 块);当最佳展示形式依赖浏览器交互且用户并未要求写入 Obsidian 时,也可以输出独立 .html / .svg。**务必在用户讨论概念、学习路线、系统架构、流程步骤、对比分析、数据关系等场景时主动触发本 skill**,哪怕用户没有明说"画一下""可视化""做个图"——只要内容具备空间结构、时序流程、分类对比或可交互演示特征,视觉化就能显著提升展示效果。凡是涉及 Obsidian 学习笔记、知识卡片、概念图、学习路径、架构梳理、流程整理、状态转移、知识对比等用户意图时,都要优先考虑使用本 skill。
Explain new concepts, technical jargon, and specialized terms at the smallest practical conceptual granularity. Use when the user is new to a topic, says a term is confusing, asks for a more detailed or step-by-step explanation, or when a concise answer would otherwise assume background knowledge the user may not have.
Generate an interactive HTML quiz on a topic the user wants to practice, review, or be tested on. Use this skill whenever the user asks to make practice questions, test their knowledge, quiz themselves, or review material — phrases like "出一份...的练习题", "考考我...", "帮我复习...", "给我做个...的quiz/小测", "测测我对...的掌握", "practice", "quiz me on", "test my knowledge". Produces a self-contained HTML file with single choice, multiple choice, true/false, and fill-in-blank questions, featuring immediate feedback, scoring, timing, and shuffling. Defaults to sourcing content from the web; only uses local markdown notes when the user explicitly asks to. Saves to the Windows Desktop and opens in the default browser.