Added Oct 16, 2026
Academic Skills for Claude Code & Codex
A Chinese-first paper-writing Skill for researchers that drafts sections, revises and answers reviewers without inventing data or references.
- Ecosystem
- Agent Skills
- Skill category
- Research
- Creator / organization
- zLanqing
What it is
Academic Skills for Claude Code & Codex is an open-source academic Skill pack by GitHub user zLanqing, written for researchers in China and usable in both Claude Code and Codex. The repository has 3 Skills: paper writing, academic Word and PowerPoint documents, and a scientific computing toolkit. We recommend only the paper-writing one, research-writing-skill. It drafts the abstract, introduction, related work, method, experiments, discussion and conclusion, revises and polishes, answers reviewers, and can turn a vague idea into a paper outline. It writes in Chinese by default and keeps English titles, formulas, variables, method names and references as they are. What sets it apart is that it labels each piece of information as one of four kinds (from the source, confirmed by you, inferred, or a suggested addition), and it's told never to invent data, DOIs, journal details or experimental results.
Who it's for:students and researchers writing papers, theses and reviewer responses in Chinese.
Before you start: copy the research-writing-skill folder into your Claude Code or Codex Skill folder. The README's install example contains a "your username" placeholder, so don't run it as written. We don't recommend the other two Skills: the Office one runs local scripts and, on Windows, drives PowerPoint; the computing toolkit's nested reference material includes advice to pipe a remote script from curl straight into bash, a script that uses free proxies to get around Google Scholar rate limits, and medical review content. The repository root is MIT-licensed, but the nested PDF and XLSX folders carry Anthropic's proprietary license and can't be used or redistributed as MIT. Your paper is processed by the AI you use, at its usual cost.
Based on the project's public documentation. Not tested by AI Inspiration World.