noi1r/beamer-skill 是从 https://github.com/noi1r/beamer-skill. 导入的特工技能候选人。
概览
noi1r/beamer-skill 是从 https://github.com/noi1r/beamer-skill. 导入的特工技能候选人。
概述（AI 翻译）
一个用于创建、编译、审核和润色学术 Beamer LaTeX 演示文稿的 AI 编码助手技能。
支持 Claude Code、OpenAI Codex CLI、Google Antigravity 和 VS Code AI 扩展（Copilot、Cline、Cursor）。
完整生命周期：创建→编译→审查→润色→验证。
需要一个可编辑的PPT吗？请使用下面的可选先用Beamer，后用PowerPoint的流程：在Beamer中建立论点、技术深度和视觉结构，然后将经过验证的.tex和.pdf交给Codex Presentations，进行可编辑的.pptx重建。
特色
| 动作 | 描述 |
|---|---|
create [topic] |
协作、迭代讲座创建，带有相位门（材料分析→需要访谈→结构计划→草稿→质量循环） |
compile [file] |
3通XeLaTeX + bitex带编译后诊断 |
review [file] |
只读校对报告（语法、拼写错误、溢出、一致性、学术质量） |
audit [file] |
可视化版面审核（溢出、字体、框、间距） |
pedagogy [file] |
包含13种验证模式的整体教学法综述 |
本页由 AI 翻译，技术细节请以原文为准。
README
以下为原文 README 节选，完整内容请查看原文。
Beamer Skill
An AI coding assistant skill for creating, compiling, reviewing, and polishing academic Beamer LaTeX presentations.
Supports Claude Code, OpenAI Codex CLI, Google Antigravity, and VS Code AI extensions (Copilot, Cline, Cursor).
Full lifecycle: create → compile → review → polish → verify.
Need an editable PowerPoint at the end? Use the optional Beamer-first, PowerPoint-second workflow below: establish the argument, technical depth, and visual structure in Beamer, then hand the verified .tex and .pdf to Codex Presentations for an editable .pptx rebuild.
Features
| Action | Description |
|---|---|
create [topic] |
Collaborative, iterative lecture creation with phase gates (material analysis → needs interview → structure plan → draft → quality loop) |
compile [file] |
3-pass XeLaTeX + bibtex with post-compile diagnostics |
review [file] |
Read-only proofreading report (grammar, typos, overflow, consistency, academic quality) |
audit [file] |
Visual layout audit (overflow, fonts, boxes, spacing) |
pedagogy [file] |
Holistic pedagogical review with 13 validation patterns |
tikz [file] |
TikZ diagram review and SVG extraction |
excellence [file] |
Comprehensive multi-dimensional quality review |
devils-advocate [file] |
Challenge slide design with pedagogical questions |
visual-check [file] |
PDF-based visual verification of compiled slides |
validate [file] |
Structural validation against skill constraints |
extract-figures [pdf] |
Extract figures from paper PDFs for direct inclusion in slides |
Highlights
- Quality scoring system — automated rubric (start at 100, deduct per issue) with Excellent/Good/Needs Work/Poor thresholds
- No overlays — no
\pause,\onslide,\only. Uses multiple slides and color emphasis instead - Content density guards — upper bounds (7 bullets, 2 equations, 5 symbols per slide) and lower bounds (every slide earns its place)
- Box overflow detection — Beamer suppresses overflow warnings inside blocks; the skill catches them via visual audit
- Motivation before formalism — every concept starts with “Why?” before “What?”
- TikZ precision — mathematical accuracy enforced via
\pgfmathsetmacro(no hardcoded approximations) - Semantic color system — colorblind-safe palette (
\pos{},\con{},\HL{}) with WCAG AA contrast (≥ 4.5:1) - Figure extraction — pull figures directly from paper PDFs via
pdf-mcp, ready for\includegraphics - Markdown paper input — use
.mdas a clean intermediate when PDF parsing loses structure; keep the PDF available for figures - Timing allocation — built-in slide-count heuristics for 5min lightning talks through 90min lectures
- Columns & layout rules — enforced
columns[T]patterns with gap/width constraints - Backup slides — automatic appendix section for anticipated Q&A
推荐工具
换一个关键词，或者移除筛选条件。
安装
npx skillfish add noi1r/beamer-skill