gitmyhub

ieee-skills

Python ★ 174 updated 1d ago

面向 IEEE 风格论文写作、润色、审稿、实验、图表、LaTeX、引用和论文阅读的 Codex skills。

A set of Codex skills that help researchers draft, polish, and pre-review IEEE-style engineering papers before submission.

Codex skillsPowerShellBashPythonLaTeX/IEEEtransetup: easycomplexity 2/5

ieee-skills is a set of nine Codex skills built for people writing, polishing, and reviewing IEEE-style engineering papers, covering conferences, journals, Transactions, and Letters. It is not an official IEEE project. Instead it packages the recurring demands of IEEE review, such as object, operating condition, engineering hazard, method rationale, and experimental evidence, into reusable workflows a writer can call on.

The skills split into three stages. The first stage covers drafting and reviewer simulation: ieee-writing helps build titles, abstracts, introductions, related work, methods, and conclusions; ieee-polishing rewrites text into IEEE-style English, including converting from Chinese; ieee-reviewer simulates an IEEE reviewer checking scope, novelty, validity, data, clarity, and compliance. The second stage strengthens evidence: ieee-experiment builds a claim-evidence matrix and flags missing experiments, baselines, or robustness checks; ieee-latex handles IEEEtran compilation, floats, equations, tables, and BibTeX problems; ieee-response drafts point-by-point responses and cover letters for revisions. The third stage covers presentation: ieee-figure-table audits whether charts and captions hold up in a two-column layout, ieee-citation checks whether references actually support the related work claims, and ieee-paper-reader extracts contributions, methods, and limitations from published IEEE papers.

The repo ships three zero-dependency SVG example figures generated by a Python script, demonstrating a robustness curve, an accuracy-latency plot, and an ablation results table, aimed at fields like AI, communications, control, signal processing, power electronics, embedded systems, and robotics.

Installation is a git clone followed by a PowerShell script on Windows or a bash script on macOS, Linux, or Git Bash, which copies the skills into a Codex skills folder (defaulting to a hidden folder in the user's home directory, with an option to choose another destination). Each skill lives in its own folder under skills/, alongside a shared references folder that the documentation says should not be skipped. The project is MIT licensed and points readers to official IEEE author resources for anything the templates leave uncertain.

Where it fits