Summary

This repository provides an AI-powered toolkit for post-writing academic paper enhancement. It addresses common issues like formatting, grammar, style, and submission readiness, offering features such as de-AI editing, reference checking, and reviewer-style audits. Designed for researchers and academics, it aims to significantly improve the quality and compliance of existing academic text across LaTeX, Typst, and PDF formats.

How It Works

The toolkit operates as a collection of specialized "skills" designed for integration with AI environments like Claude Code. Each skill leverages large language models and custom Python scripts to perform targeted analyses and improvements. Core functionalities include format validation, grammar and expression polishing, logical coherence checks (using the AXES model), literature review synthesis, de-AI editing to humanize AI-generated text, and comprehensive paper audits. This modular, AI-driven approach allows for flexible and context-aware academic writing assistance.

Quick Start & Requirements

Installation is facilitated via Claude Code's skills manager: npx skills add bahayonghang/academic-writing-skills. Key prerequisites include Python 3.10+, a LaTeX distribution (TeX Live/MiKTeX with latexmk, chktex), or Typst CLI. Chinese language support requires specific CJK fonts. PDF analysis for the paper-audit skill necessitates pdfplumber. Users will also need an AI environment capable of running these skills.

Highlighted Details

• Comprehensive Skills: Offers specialized modules for English LaTeX papers (latex-paper-en), Chinese theses (latex-thesis-zh), Typst papers (typst-paper), bibliography management (bib-search-citation), deep paper audits (paper-audit), and industrial AI literature synthesis (industrial-ai-research).
• De-AI Editing: Intelligently humanizes AI-generated passages while preserving all code and LaTeX/Typst syntax.
• Advanced Auditing: The paper-audit skill provides reviewer-style critiques, ScholarEval scoring, and submission gate checks for .tex, .typ, and .pdf files.
• Bibliography Verification: Features online verification against CrossRef and Semantic Scholar without requiring API keys.
• Format & Venue Support: Accommodates major academic venues (IEEE, ACM, Springer, NeurIPS) and Chinese thesis standards (GB/T 7714-2015), supporting both LaTeX and Typst.

Maintenance & Community

The project welcomes issues and pull requests, encouraging contributions scoped to relevant skills. Specific details regarding active maintainers, community channels (like Discord/Slack), or sponsorships are not detailed in the README.

Licensing & Compatibility

The software is licensed strictly for "Academic Use Only" and is explicitly prohibited from commercial use. This restriction impacts its compatibility with closed-source or commercial projects that require open-source components.

Limitations & Caveats

The primary limitation is the strict "Academic Use Only" license, prohibiting any commercial application. Setup requires significant technical prerequisites, including specific LaTeX or Typst installations and an AI-powered execution environment. PDF analysis in the paper-audit skill is limited to text-only checks. Debugging compilation errors necessitates user-provided log snippets.