English | 中文 | 日本語 | Français | Español | العربية | 한국어 | Português | Русский | Deutsch

🧪 Awesome Skills

AI Coding Assistant Skills for Researchers

🎯 Stop explaining your research workflow to AI every time.
Write the skill once. Get perfect results forever.

---

💡 What is this?

Awesome Skills is a curated collection of AI coding assistant skills — reusable instruction files that teach your AI assistant how to perform research tasks exactly the way you want.

Instead of writing the same prompt over and over, you write a skill once — and your AI assistant follows it perfectly every single time.

---

😤 Before vs After

❌ Without Skills	✅ With Skills
You: "Help me write a LaTeX paper" AI: Generates generic LaTeX template Missing \usepackage for your field No journal-specific formatting Citations in wrong format Figures not properly referenced You spend 30 min fixing everything...	You: "Write my paper using latex-paper skill" AI: Uses your target journal's exact template Correct citation format (APA/IEEE/Nature) Proper \label and \ref for all figures Statistical tests reported correctly Code compiles without errors on first try Ready to submit in 5 min ✨
You: "Analyze this CSV data" AI: Runs basic describe() and a bar chart No normality test, no effect size Missing confidence intervals Wrong statistical test for your data type You redo the entire analysis manually	You: "Analyze using statistical-analysis skill" AI: Checks data distribution first Selects correct test (parametric vs non-parametric) Reports effect sizes and CIs Generates publication-quality figures Outputs APA-formatted results section

---

🚀 Quick Start

Skills are just instruction files that tell your AI assistant how to behave. Drop them into your project and your AI will follow the instructions automatically.

Cursor (.cursorrules)

# Copy skill to your project root
cp skills/latex-paper.md .cursorrules

# Or append to existing rules
cat skills/latex-paper.md >> .cursorrules

Claude Code (CLAUDE.md)

# Copy skill to your project root
cp skills/latex-paper.md CLAUDE.md

# Or append to existing instructions
cat skills/latex-paper.md >> CLAUDE.md

Kimi Code (AGENTS.md)

# Copy skill to your project root
cp skills/latex-paper.md AGENTS.md

# Or append to existing instructions
cat skills/latex-paper.md >> AGENTS.md

💡 Pro tip: You can combine multiple skills by concatenating them into a single file!

---

📚 Skills Catalog

📝 Academic Writing / 学术写作

Skill	Description	Use When
latex-paper	LaTeX paper writing with journal-specific formatting	Writing journal or conference papers
research-proposal	Grant and thesis proposal writing	Applying for funding or writing thesis proposals
literature-review	Systematic literature review and synthesis	Reviewing a research field or writing related work
academic-english	Academic English polishing and grammar	Polishing non-native English writing

📊 Data Analysis / 数据分析

Skill	Description	Use When
statistical-analysis	Statistical testing with proper assumptions checks	Analyzing experimental data
data-visualization	Publication-quality figures (matplotlib/seaborn/plotly)	Creating figures for papers
ml-experiment	ML experiment tracking with proper baselines	Running machine learning experiments

📖 Literature Management / 文献管理

Skill	Description	Use When
zotero-integration	Zotero bibliography integration	Managing references with Zotero
citation-management	Citation formatting across styles	Formatting references for different journals

🔬 Experiment Tools / 实验工具

Skill	Description	Use When
jupyter-notebook	Clean Jupyter notebook workflows	Running and organizing notebooks
docker-reproducibility	Reproducible research environments	Setting up reproducible experiments

📬 Paper Submission / 论文投稿

Skill	Description	Use When
conference-submission	Conference submission checklist and formatting	Submitting to NeurIPS, ICML, CVPR, etc.

---

🎬 Usage Examples

Example 1: Writing a Full Paper

# 1. Load the latex-paper skill
cp skills/latex-paper.md CLAUDE.md

# 2. Ask your AI assistant
"Write a LaTeX paper for IEEE TPAMI on my transformer attention mechanism.
Include: abstract, introduction, related work, method, experiments, conclusion.
Use IEEE citation format. Reference figures as Fig. 1, Fig. 2, etc."

Example 2: Analyzing Experiment Data

# 1. Load the statistical-analysis skill
cp skills/statistical-analysis.md .cursorrules

# 2. Ask your AI assistant
"Analyze the data in results.csv. Compare model A vs model B performance.
Check normality, run appropriate tests, report effect sizes and CIs.
Generate a box plot for Figure 1."

Example 3: Complete Research Workflow

See examples/research-workflow.md for a full end-to-end workflow combining multiple skills.

---

❓ FAQ

Q: What AI assistants do these skills work with?

These skills work with any AI coding assistant that supports custom instructions:

• Cursor via .cursorrules
• Claude Code via CLAUDE.md
• Kimi Code via AGENTS.md
• Windsurf via .windsurfrules
• Cline via .clinerules
• Aider via .aider.conf.yml
• Any assistant that reads instruction files

Q: Can I combine multiple skills?

Yes! Simply concatenate the skill files:

cat skills/latex-paper.md skills/data-visualization.md > CLAUDE.md

Or use them selectively based on your current task.

Q: How are these different from just writing prompts?

Skills are:

• Persistent — saved in your project, not lost between sessions
• Version-controlled — track changes with git
• Reusable — share with your team
• Composable — combine multiple skills
• Tested — refined through real research workflows

Q: Can I modify the skills for my field?

Absolutely! Skills are just markdown files. Fork this repo and customize them for your specific research area, journal requirements, or lab conventions.

Q: Do these skills work for non-English papers?

Yes! Skills like latex-paper and academic-english can be configured for Chinese (中文), Japanese (日本語), and other languages. See each skill's documentation for language options.

---

🔗 See Also

• 🤖 awesome-ai-rules — Curated collection of AI coding assistant rules and configurations
• ✅ vibe-check — Validate your AI assistant's output quality
• 🔄 ai-commit (commit-ai) — AI-powered conventional commit messages
• 🔌 awesome-mcp-servers — Model Context Protocol server collection

---

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

Ways to contribute:

• 🆕 Add a new skill for your research area
• 🐛 Fix bugs or improve existing skills
• 📖 Improve documentation
• 💡 Suggest new skill ideas via Issues

---

📄 License

MIT © liangzhengtao

---

⬆ Back to top

---