[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-76151":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":12,"contributorsCount":12,"subscribersCount":12,"size":12,"stars1d":14,"stars7d":14,"stars30d":15,"stars90d":12,"forks30d":12,"starsTrendScore":16,"compositeScore":12,"rankGlobal":9,"rankLanguage":9,"license":17,"archived":18,"fork":18,"defaultBranch":19,"hasWiki":20,"hasPages":18,"topics":21,"createdAt":9,"pushedAt":9,"updatedAt":30,"readmeContent":31,"aiSummary":32,"trendingCount":12,"starSnapshotCount":12,"syncStatus":33,"lastSyncTime":34,"discoverSource":35},76151,"awesome-latex-skills","Calix-L\u002Fawesome-latex-skills","Calix-L","Prompt packs that make any AI agent a LaTeX expert — fix errors, polish writing, format for venues, read papers, recover source",null,"Shell",172,0,123,3,42,9,"MIT License",false,"main",true,[22,23,24,25,26,27,28,29],"academic-writing","ai-agent","claude-code","latex","llm","neurips","paper-writing","prompt-engineering","2026-06-12 02:03:40","\u003Cdiv align=\"center\">\n\n[中文文档](.\u002FREADME_CN.md)\n\n\u003Cimg src=\".\u002Fassets\u002Fbanner.svg\" alt=\"awesome-latex-skills\" width=\"100%\">\n\n\u003Cbr>\n\n### Every researcher knows the feeling.\n\n> *47 compilation errors at 2 AM. A reviewer who writes \"English needs improvement.\"*\n> *A CVPR reject that needs to become an ICML submission by Friday.*\n\n**awesome-latex-skills** turns any AI agent into a LaTeX expert — structured workflows, curated knowledge, and guardrails that raw prompts can't replicate.\n\n\u003Cbr>\n\n`47 errors → 0`  ·  `Chinglish → publication-ready`  ·  `CVPR → NeurIPS`  ·  `50 pages → structured notes`\n\n\u003Cbr>\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FCalix-L\u002Fawesome-latex-skills\u002Factions\">\u003Cimg src=\"https:\u002F\u002Fgithub.com\u002FCalix-L\u002Fawesome-latex-skills\u002Factions\u002Fworkflows\u002Ftest.yml\u002Fbadge.svg\" alt=\"CI\">\u003C\u002Fa>\n\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fskills-5-blue\" alt=\"5 skills\">\n\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Ftests-168_passing-brightgreen\" alt=\"168 tests\">\n\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FCalix-L\u002Fawesome-latex-skills?style=social\" alt=\"Stars\">\n\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-MIT-yellow\" alt=\"MIT\">\n\n\u003C\u002Fdiv>\n\n---\n\n**[10-Second Pitch](#10-second-pitch) · [Skills](#skills) · [Demos](#demos) · [Workflows](#workflows) · [Quick Start](#quick-start) · [Compatibility](#compatibility)**\n\n---\n\n## 10-Second Pitch\n\nYou describe the problem. The skill produces the fix.\n\n- **47 compilation errors at 2 AM** — `latex-rescue` auto-corrects typos, fixes mismatched environments, resolves package conflicts. 80+ patterns, zero manual edits.\n- **\"English needs improvement\" from Reviewer #2** — `latex-polish` fixes 18 categories of Chinglish, applies 100+ academic phrasebank templates, adds proper hedging. 3 intensity levels.\n- **CVPR rejected, ICML deadline Friday** — `latex-fmt` switches `\\documentclass`, removes banned packages, anonymizes, checks page limits. 15 venues covered.\n- **50 papers in your reading list** — `paper-read` produces a 5-bullet skim in 30 seconds, a structured analysis in 5 minutes, or a full critical review in 15.\n- **Lost the .tex, only the PDF** — `pdf2tex` rebuilds LaTeX from any compiled PDF. 97+ math glyph mappings, table reconstruction, 7-phase pipeline.\n\nNo LaTeX expertise required. The skill handles the semicolons.\n\n---\n\n## Skills\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"54\" align=\"center\">:ambulance:\u003C\u002Ftd>\n\u003Ctd width=\"120\">\u003Cstrong>\u003Ca href=\".\u002Flatex-rescue\u002FSKILL.md\">latex-rescue\u003C\u002Fa>\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>Fix compilation errors — 80+ auto-fix patterns, package conflicts, Overleaf support\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\">:pencil2:\u003C\u002Ftd>\n\u003Ctd>\u003Cstrong>\u003Ca href=\".\u002Flatex-polish\u002FSKILL.md\">latex-polish\u003C\u002Fa>\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>Polish academic writing — 18 Chinglish categories, 100+ phrasebank templates, 3 intensity levels\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\">:repeat:\u003C\u002Ftd>\n\u003Ctd>\u003Cstrong>\u003Ca href=\".\u002Flatex-fmt\u002FSKILL.md\">latex-fmt\u003C\u002Fa>\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>Reformat between 15 venues — NeurIPS · ICML · CVPR · ACL · ICLR · ECCV · AAAI · TMLR · IEEE · Nature · Science · COLING · KDD · SIGIR · Interspeech\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\">:book:\u003C\u002Ftd>\n\u003Ctd>\u003Cstrong>\u003Ca href=\".\u002Fpaper-read\u002FSKILL.md\">paper-read\u003C\u002Fa>\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>Read & analyze papers — skim \u002F read \u002F deep, 50+ critical appraisal items, assumption auditing\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\">:wrench:\u003C\u002Ftd>\n\u003Ctd>\u003Cstrong>\u003Ca href=\".\u002Fpdf2tex\u002FSKILL.md\">pdf2tex\u003C\u002Fa>\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>Rebuild LaTeX from PDF — 7-phase pipeline, 97+ math glyph mappings, table reconstruction\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n| | | | |\n|---|---|---|---|\n| 80+ error patterns | 14 package conflicts | 18 Chinglish categories | 100+ phrasebank templates |\n| 15 venue rules | 50+ appraisal items | 97+ glyph mappings | 15 reference files |\n\n---\n\n## Why skills, not just prompts?\n\nYou've tried asking ChatGPT to fix your LaTeX. It guesses. It misses things. It changes your math.\n\n| You say... | What the raw LLM does | What the skill pack does |\n|---|---|---|\n| `\\beginn{table}` | \"That's an interesting typo\" | Auto-corrects to `\\begin{table}` |\n| \"According to the experiment\" | Accepts it | Flags overuse, suggests alternatives |\n| \"Format for NeurIPS\" | Forgets Broader Impact | Flags missing required section |\n| \"Convert this PDF to LaTeX\" | Produces broken markup | 7-phase pipeline with verification |\n| `\\citep{}` without natbib | Silently ignores | Detects missing package, adds it |\n| \"Polish my paper\" | Rewrites everything | Minimal edits, preserves math & commands |\n\nSkills inject **hundreds of domain-specific rules** that LLMs can't reliably recall from memory. Each skill = structured workflow + reference knowledge + guardrails. Same input, same expert output, every time.\n\n---\n\n## Demos\n\n### :ambulance: latex-rescue — *2 AM, 47 errors, deadline tomorrow*\n\n```diff\n- \\textbff{bold}              → Undefined control sequence\n+ \\textbf{bold}               → auto-fixed\n\n- x_i is important            → Missing $ inserted\n+ $x_i$ is important          → auto-fixed\n\n- \\begin{figure}...\\end{table}\n+ \\begin{figure}...\\end{figure}  → mismatch fixed\n```\n\n### :pencil2: latex-polish — *Reviewer #2 says \"English needs improvement\"*\n\n```diff\n- The model can achieves good performance on the dataset.\n+ The model achieves strong performance on the benchmark.\n\n- According to the experiment, it makes the accuracy improved by 3.2%.\n+ Experiments show that the method improves accuracy by 3.2%.\n\n- Most of methods in this research field can not achieve the same result.\n+ Most methods in this field fail to match this result.\n```\n\n### :repeat: latex-fmt — *Camera-ready reformat, CVPR → NeurIPS*\n\n```diff\n- \\documentclass{article}\n+ \\documentclass{neurips_2025}\n- \\author{Zhang et al.}\n+ \\author{Anonymous}\n- (no Broader Impact section)\n+ ⚠ Broader Impact required by NeurIPS — flagged\n```\n\n### :book: paper-read — *50 papers in your reading list, no time*\n\n```diff\n- \"This paper proposes a novel transformer-based approach for...\"\n+ [skim] Object detection · Wang et al., CVPR 2024\n+        Novelty: sparse attention for real-time. Verdict: worth deep read.\n\n- (reading every paper front-to-back)\n+ [deep] Key eq: sparse attention. Delta: 10x faster.\n+        Gap: only tested on COCO. Overclaim: \"SOTA\" (margin 0.3%).\n```\n\n### :wrench: pdf2tex — *Lost the .tex, only the PDF survives*\n\n```diff\n- (staring at a compiled PDF, no source files)\n+ \\documentclass{article}\n+ \\usepackage{amsmath,amssymb}\n+ \\section{Introduction}\n+ The model achieves $F_1 = 92.3$ on the benchmark.\n+ % [UNCERTAIN: math notation — verify subscripts]\n```\n\n---\n\n## Workflows\n\nSkills compose into pipelines for real academic scenarios:\n\n| Scenario | What you type | What happens |\n|---|---|---|\n| Deadline crunch | `\u002Flatex-rescue` | Crash → rescue → compile |\n| Review turnaround | `\u002Flatex-polish` → `\u002Flatex-fmt` | Draft → polish → format → submit |\n| Rebuttal reformat | `\u002Flatex-polish` → `\u002Flatex-fmt` | CVPR reject → polish → reformat for ICML |\n| Lost source | `\u002Fpdf2tex` → `\u002Flatex-rescue` | PDF → reconstruct → fix → compile |\n| New paper | `\u002Fpaper-read` → `\u002Flatex-polish` → `\u002Flatex-fmt` | Read papers → polish → format for venue |\n| Overleaf | `\u002Flatex-rescue` | Paste error log → get fixes |\n\n---\n\n## Quick Start\n\n**One command to install all 5 skills:**\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FCalix-L\u002Fawesome-latex-skills.git && \\\ncp -r awesome-latex-skills\u002F{latex-rescue,latex-polish,latex-fmt,paper-read,pdf2tex} ~\u002F.claude\u002Fskills\u002F\n```\n\nThen just type `\u002Flatex-rescue`, `\u002Flatex-polish`, etc. in Claude Code.\n\n**One skill only:**\n```bash\ncp -r awesome-latex-skills\u002Flatex-rescue ~\u002F.claude\u002Fskills\u002F\n```\n\n**Not using Claude Code?** Just point your agent to the SKILL.md:\n```\nRead awesome-latex-skills\u002Flatex-rescue\u002FSKILL.md and follow the workflow.\n```\n\n\u003Csup>No install needed for `latex-polish`, `latex-fmt`, or `paper-read`. `latex-rescue` needs LaTeX. `pdf2tex` needs `pip install pymupdf`.\u003C\u002Fsup>\n\n---\n\n## Compatibility\n\n| Platform | How to use |\n|---|---|\n| **Claude Code** | Copy to `~\u002F.claude\u002Fskills\u002F`, invoke with `\u002Flatex-rescue` |\n| **ChatGPT \u002F GPT-4** | Paste SKILL.md as custom instruction or system prompt |\n| **Cursor** | Add SKILL.md content to `.cursor\u002Frules\u002F` |\n| **Copilot** | Add SKILL.md content to `.github\u002Fcopilot-instructions.md` |\n| **Any LLM** | Send SKILL.md as context, then ask your question |\n\n---\n\n\u003Cdetails>\n\u003Csummary>How it works\u003C\u002Fsummary>\n\nEach skill is a self-contained directory:\n\n```\nlatex-rescue\u002F\n├── SKILL.md              # the prompt — role, triggers, workflow, guardrails\n├── references\u002F           # domain knowledge the agent reads at each phase\n│   ├── error-catalog.md\n│   ├── package-conflicts.md\n│   └── debug-workflow.md\n└── agents\u002F\n    └── config.yaml       # auto-activation triggers and platform settings\n```\n\n1. You type `\u002Flatex-rescue` or say \"fix my LaTeX errors\"\n2. Agent loads `SKILL.md` — now it has a structured workflow + guardrails\n3. It reads `references\u002F` for precise domain rules at each phase\n4. Same workflow → same expert output, every time\n\n\u003C\u002Fdetails>\n\n---\n\n## License\n\nMIT","awesome-latex-skills 是一个旨在将任何AI代理转变为LaTeX专家的项目，通过提供结构化的工作流程、精选的知识库和防护措施来解决LaTeX文档编译错误、改进学术写作质量、格式调整以及论文阅读与源码恢复等问题。其核心功能包括自动修复常见的LaTeX编译错误（如错别字、环境不匹配等）、提升英文表达至出版水平、快速转换文档格式以适应不同学术会议要求、从PDF文件中恢复LaTeX源代码等，且无需用户具备深厚的LaTeX知识。适用于科研人员在撰写或修改学术论文时遇到的各种挑战场景，尤其是对于那些需要频繁调整格式以符合不同期刊或会议要求的研究者来说非常实用。",2,"2026-06-11 03:54:39","CREATED_QUERY"]