zlanqing/codex-claude-academic-skills

This repository contains three companion skills built for Chinese academic researchers, designed to be used inside Claude Code and Codex, which are AI coding assistant tools. Each skill targets a different part of the research workflow, and they are meant to work together on the same project. The first skill, research-writing-skill, helps with writing academic papers: drafting sections such as the abstract, introduction, related work, methods, experiments, discussion, and conclusion, as well as revising and polishing existing text for logical consistency and correct terminology. It can also help draft responses to peer reviewer comments and turn a rough research idea into a paper outline. The writing rules favor Chinese as the default language while keeping English terms, formulas, variable names, method names, and citations unchanged, and the skill is careful not to invent data, identifiers, or journal information that was not already provided. The second skill, office-academic-skill, focuses on generating and editing academic Word documents and PowerPoint slides, aimed at tasks like turning a literature review into a formatted report, or building slide decks for group meetings, thesis proposals, and thesis defenses. The third skill, scientific-toolkit-skill, combines MATLAB and Python for scientific computing tasks such as simulations, signal processing, statistics, and machine learning, along with generating publication style figures for papers. The three skills are designed to complement each other. In a typical paper writing workflow, for example, the toolkit skill produces the data analysis and figures, the writing skill turns that into the paper's text, and the office skill formats the final defense slides. The project is released under the MIT license and is written in Python.