latex-thesis-zh_skill

This skill helps you refine Chinese LaTeX thesis writing, ensuring structure, style, and formatting while preserving citations and math.
  • Python

10

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill bahayonghang/my-claude-code-settings --skill latex-thesis-zh

  • README.md6.0 KB
  • SKILL.md9.6 KB

Overview

This skill is a Chinese LaTeX thesis assistant tailored for doctoral and master's theses in deep learning, time series, and industrial control. It provides modular tools for compilation, structure mapping, GB/T format checks, academic polishing, logical coherence, long-sentence analysis, bibliography validation, template detection, de-AIfication, and title optimization. The skill enforces conservative editing rules to preserve citations, labels, and math environments and outputs change suggestions as annotated LaTeX comments. It is designed for iterative, auditable edits and integrates optional scripts for automated checks and compilation when explicitly requested.

How this skill works

You invoke specific modules via trigger words (e.g., "compile", "structure", "format", "expression", "logic", "long sentence", "bib", "template", "deai", "title"). Each module inspects the LaTeX source and returns annotated suggestions in a fixed comment template that includes severity and priority. Compilation routines use latexmk with XeLaTeX/LuaLaTeX heuristics for Chinese documents; analysis modules parse source to avoid touching \cite{}, \ref{}, \label{}, and math environments. Destructive actions (clean, full rebuild) and script execution only run when you explicitly confirm.

When to use it

  • Prepare or proof a Chinese thesis in deep learning, time series, or industrial control.
  • Need safe automated compilation and dependency handling for XeLaTeX/LuaLaTeX workflows.
  • Validate compliance with GB/T bibliographic and formatting conventions.
  • Polish academic phrasing, improve paragraph-level logic, or reduce AI-style wording.
  • Generate or optimize thesis titles to maximize clarity and discoverability.

Best practices

  • Always provide the main .tex path and target chapter when requesting edits; the tool will ask if arguments are missing.
  • Run structure mapping before full-document reviews or when multiple files are involved.
  • Request compilation or destructive clean operations only after reviewing suggested commands and confirming intent.
  • Supply source snippets (preserve indentation) for expression, de-AIfication, or long-sentence modules to get minimal-intrusion edits.
  • Treat all outputs as annotated suggestions; apply changes to source yourself or authorize the script to write files.

Example use cases

  • Compile a Chinese thesis with latexmk + XeLaTeX, auto-running biber/bibtex as needed and returning first-error diagnostics.
  • Map multi-file thesis structure and detect the applied university template and missing front/back matter.
  • Check references.bib for missing required fields, duplicates, and GB/T 7714 compliance while flagging unused or missing cites.
  • Polish a paragraph from colloquial Chinese to formal academic Chinese while preserving citations and math.
  • Analyze a long, 90-character sentence and propose a split with a clear main clause and supporting phrases.

FAQ

No. The skill never alters citations, labels, or content inside math environments; all suggestions are emitted as commented diffs.

Can the skill run compilation and clean commands automatically?

Yes, but only when you explicitly request compilation or destructive clean options; the skill will present the exact command and ask for confirmation before execution.

How does de-AIfication work without changing facts or introducing new data?

De-AIfication rewrites visible prose for naturalness, removes empty buzzwords, and annotates claims that need evidence—without adding facts, figures, or new citations.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational
latex-thesis-zh skill by bahayonghang/my-claude-code-settings | VeilStrat