为什么值得关注

它采用独特的“织布”隐喻来组织笔记，内置填空、温度计等命令，并提供一个 AI 技能，能从章节文本自动生成完整的 Loom 笔记本。

适合谁使用

学习数学或技术课程的学生
希望制作主动学习讲义的教育工作者
偏好纸质主动回忆的自学者
对染料主题美学排版感兴趣的 LaTeX 用户

典型使用场景

为线性代数、信号处理或优化课程创建填空学习指南
为定理密集的课程生成交互式证明框架
制作带有织布封面和颜色编码元素的精美笔记
使用 AI 技能将教科书章节转换为可直接编译的 Loom 笔记本

项目优势

通过 \fillin、\TODO 和 yourturn 环境，在同一文档中同时支持被动阅读和主动填空练习
提供温度计 (\warmth) 和边距回忆提示 (\recall) 来跟踪理解程度
包含一个 AI 技能，能从原始材料自动生成 Loom 笔记本
提供兼容 Overleaf 的版本和包含完整命令参考的模板

使用前须知

需要 XeLaTeX 和特定的 macOS 字体（Optima、宋体、楷体）以获得完整视觉效果；非 macOS 用户必须修改字体配置
AI 技能专为 Claude 设计，可能无法直接在其他 AI 工具上使用
示例笔记基于公有领域数学或跟随受版权保护的文本；用户需确保自己内容的正确署名

README 快速开始

🧶 Loom

Weave your understanding — a gorgeous XeLaTeX class for fill-in study notes

Notes you both read and fill: statements and intuition to take in, blanks and proof-skeletons to work out. Learn by active recall, on paper that looks like an illuminated manuscript.

Why Loom

Most notes are read-only: definition → theorem → proof, flowing past your eyes. You finish the page and remember nothing. Loom typesets the other layer too — the weave of your own understanding:

🧵 A page is cloth on a loom. Warp threads are the ideas in play (a left selvage rail). Knots are results that bind them (theorem boxes). Loose threads are your open questions. A warmth gauge records how well you actually grok each block.
✍️ Passive + active in one document. The exposition is written to be read; the high-value steps are left blank (\fillin), proofs ship as skeletons (\TODO), and the book's examples are restaged as "your turn" computations. The source is your answer key.
🎨 Dyed in real pigments. Indigo, madder, weld, iron-gall — historical textile and ink dyes no maths template uses. Headings incised in Optima, body and math in Libertinus, 中文 in Songti/楷体. Every cover carries a procedurally-woven emblem.
🤖 An AI skill that writes them for you. Point Claude at a chapter and the fill-in-notes skill produces a complete, compiling notebook in this style — organized around a spine, not transcribed.

Singular Value Decomposition — a definition knot, the geometry, a Your turn Fourier transform — the dictionary, the convolution theorem

Entropy — the information diagram, Gibbs' inequality Gradient descent — the rate, and the cost of conditioning

Concentration of measure — theorem stated, proof left as a skeleton Every device on one page (the demo)

Quickstart

git clone https://github.com/Polaris-Aeterna/loom-notes.git
cd loom-notes/template
latexmk -xelatex main.tex      # or: xelatex main.tex (twice)

Then write. A blank starter, doubling as a live cheat-sheet of every command, is in template/main.tex. Open the repo in VS Code and the bundled .vscode/settings.json builds with XeLaTeX on every save (LaTeX Workshop defaults to pdflatex, w

项目描述

A gorgeous XeLaTeX class + Claude skill for fill-in study notes — read and fill, learn by active recall.

loom-notes