Claude Code Plugins

Community-maintained marketplace

Feedback

|

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name latex-thesis-zh
description Chinese doctoral/master thesis LaTeX assistant. Provides: structure integrity checking, GB/T 7714 format validation, compilation (xelatex/lualatex via latexmk), Chinese academic expression standards, terminology consistency analysis. Use when: - Working with Chinese thesis projects (main.tex + chapter files) - User mentions "doctoral thesis", "master thesis", "dissertation", "graduation thesis" - User mentions Chinese: "博士论文", "硕士论文", "学位论文", "毕业论文" - User asks about university templates: thuthesis, pkuthss, ustcthesis, fduthesis - User needs GB/T 7714 bibliography format checking - User wants to compile Chinese LaTeX documents

LaTeX Thesis Assistant (Chinese / 中文博士论文)

Critical Rules / 核心原则

绝对禁止违反以下规则:

  1. 绝不修改 \cite{}\ref{}\label{}、公式环境内的内容
  2. 绝不凭空捏造参考文献条目 - 只能读取现有 .bib 文件
  3. 绝不在未经许可的情况下修改专业术语(见 FORBIDDEN_TERMS.md
  4. 绝不直接应用修改 - 始终先以注释形式输出 diff
  5. 必须先运行结构映射,再开始分析
  6. 中文文档必须使用 XeLaTeX 或 LuaLaTeX 编译

Quick Start / 快速开始

Compile Document / 编译文档

# Auto-detect and compile (XeLaTeX for Chinese)
python scripts/compile.py main.tex

# Explicit compiler (XeLaTeX recommended for Chinese)
python scripts/compile.py main.tex --compiler xelatex
python scripts/compile.py main.tex --compiler lualatex

# Using recipes (VS Code LaTeX Workshop style) / 使用编译配方
python scripts/compile.py main.tex --recipe xelatex           # XeLaTeX only
python scripts/compile.py main.tex --recipe latexmk           # LaTeXmk auto
python scripts/compile.py main.tex --recipe xelatex-bibtex    # xelatex -> bibtex -> xelatex*2
python scripts/compile.py main.tex --recipe xelatex-biber     # xelatex -> biber -> xelatex*2 (推荐)
python scripts/compile.py main.tex --recipe pdflatex-bibtex   # pdflatex -> bibtex -> pdflatex*2
python scripts/compile.py main.tex --recipe pdflatex-biber    # pdflatex -> biber -> pdflatex*2

# Continuous compilation (watch mode)
python scripts/compile.py main.tex --watch

# Clean auxiliary files / 清理辅助文件
python scripts/compile.py main.tex --clean

Structure Mapping / 结构映射

# Map thesis structure (MUST run first)
python scripts/map_structure.py main.tex

Workflow (5-Layer) / 工作流程

Layer 0: Structure Mapping (MANDATORY)

python scripts/map_structure.py main.tex

Output: File tree + template type + processing order

Layer 1: Structure Integrity Check

Check against STRUCTURE_GUIDE.md:

Front Matter (必须):

  • Cover, Declaration, Abstract (CN+EN), TOC
  • Symbol list (recommended)

Main Body:

  • Chapter 1: Introduction (background, significance, contributions)
  • Chapter 2: Related Work
  • Chapters 3-N: Core content
  • Final Chapter: Conclusions

Back Matter (必须):

  • References, Acknowledgments, Publication list

Layer 2: GB/T 7714 Format Check

See GB_STANDARD.md:

  • Bibliography format (biblatex-gb7714-2015)
  • Figure/table captions (宋体五号)
  • Equation numbering (e.g., (3.1))
  • Section heading styles

Layer 3: Chinese Academic Expression

See ACADEMIC_STYLE_ZH.md:

Oral → Academic conversion:

  • ❌ "很多研究表明" → ✅ "大量研究表明"
  • ❌ "效果很好" → ✅ "具有显著优势"

Forbidden subjective words:

  • ❌ 显然、毫无疑问、众所周知
  • ✅ 研究表明、实验结果显示、可以认为

Layer 4: Long Sentence Decomposition

Trigger: >60 characters or >3 clauses

% 长难句检测(第45行,共87字)
% 主干:本文方法在多个数据集上取得优异性能。
% 修饰成分:
% - [定语] 基于深度学习的
% - [方式] 通过引入注意力机制
% 建议改写:[分拆版本]

Layer 5: Expression Restructuring

% ═══════════════════════════════════════════
% 修改建议(第23行)
% ═══════════════════════════════════════════
% 原文:我们使用了ResNet模型。
% 修改后:本文采用ResNet模型作为特征提取器。
% 改进点:
% 1. "我们使用" → "本文采用"(学术规范)
% 2. 补充模型用途说明
% ═══════════════════════════════════════════

Template Detection / 模板检测

Auto-detect university template:

python scripts/detect_template.py main.tex

Supported templates (see UNIVERSITIES/):

  • thuthesis: Tsinghua University
  • pkuthss: Peking University
  • ustcthesis: USTC
  • fduthesis: Fudan University
  • generic: General GB/T compliant

Output Report / 输出报告

# LaTeX 博士论文审查报告

## 总览
- 整体状态:✅ 符合要求 / ⚠️ 需要修订 / ❌ 重大问题
- 编译状态:[status]
- 模板类型:[detected template]

## 结构完整性(X/10 通过)
### ✅ 已完成项
### ⚠️ 待完善项

## 国标格式审查
### ✅ 符合项
### ❌ 不符合项

## 中文学术表达(N处建议)
[按优先级分组]

## 长难句拆解(M处)
[详细分析]

References / 参考文档