Summary

This project offers LaTeX templates for Chinese scientific research, particularly grant proposals, enhanced by an AI-driven "Vibe Writing" methodology. It targets researchers automating tedious tasks like formatting and reference management, enabling efficient, collaborative writing leveraging AI advancements.

How It Works

The core is "Vibe Coding/Writing," positioning AI as an active collaborator. AI analyzes templates, automates style alignment, and handles compilation under human oversight. A modular "Agent Skills" system integrates AI capabilities for literature review to proposal drafting, designed to evolve with LLM progress.

Quick Start & Requirements

Local development is recommended via VS Code + LaTeX Workshop. Installation involves integrating AI assistants (Claude Code/Codex CLI) using a specific prompt to install project skills. Prerequisites include LaTeX (TeX Live recommended) and XeLaTeX compiler. AI model access (GPT-5.2, Claude 4.5 Opus, GLM-4.7) is needed, with guidance for third-party API services suitable for Chinese users.

Highlighted Details

• AI-Driven Automation: Smart template planning, style alignment, code review, testing.
• Modular Skill System: Extensible AI skills cover literature review to proposal drafting.
• Human-AI Collaboration: AI handles mechanical tasks, researchers focus on creativity.
• Template Conversion: High-fidelity Word/PDF to LaTeX conversion via make_latex_model.
• NSFC Focus: Templates for Youth, General, Regional grants ("2026 Ready").

Maintenance & Community

Maintained by @huangwb8. Community interaction via WeChat group for experience sharing and GitHub Issues for bug reports/features.

Licensing & Compatibility

Released under the MIT License, permitting commercial use and integration into closed-source projects.

Limitations & Caveats

"2026 Ready" templates are speculative and subject to official NSFC announcements. Several AI skills (nsfc-bib-manager, etc.) are "🚧 Development" and may lack stability. Optimal AI access may require third-party APIs due to network constraints. Local development is strongly preferred over online platforms like Overleaf.