awesome-ai-research-writing
Elevate your AI research writing, no more tedious polishing ✨
awesome-ai-research-writing is a comprehensive collection of research writing prompts and agent skills for AI assistants used in academic paper writing. Created by surveying researchers at top institutions (MSRA, Seed, SH AI Lab, Peking University, USTC, SJTU), it provides battle-tested prompt templates for translation, polishing, logical checking, and citation management, plus installable agent skills for structured paper writing workflows. The resource targets ML/CS researchers who use AI coding tools like Cursor or Claude Code and want production-ready prompts instead of trial-and-error prompt engineering. Skills can be installed via OpenSkills and automatically integrate into the agent's workflow, enabling one-command actions like "write an ICLR paper from this repo" or "migrate this NeurIPS draft to ICML format."
Key Features
Use Cases
- 01Translating Chinese research notes to publication-ready English LaTeX manuscripts for top-tier ML conferences
- 02Generating perfectly formatted booktabs comparison tables from CSV experiment results with automatic best-value highlighting
- 03Migrating accepted papers between conference formats (e.g., NeurIPS camera-ready to JMLR submission) while preserving content
- 04Running pre-submission checklists that verify anonymization, page limits, citation formats, and ethics statements per venue
- 05Performing final-stage "AI smell" removal to eliminate formulaic transitions and overused academic buzzwords before submission
- 06Creating LaTeX-compatible framework diagrams and architecture figures with automatic caption generation following style guides
Related Skills
View moresuperpowers
An agentic skills framework & software development methodology that works.
skills
Skills for Real Engineers. Straight from my .agents directory.
skills
Public repository for Agent Skills
ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
awesome-ai-research-writing — FAQ
What is awesome-ai-research-writing?+
It's a curated collection of 50+ research writing prompts and installable agent skills designed for AI coding assistants (Cursor, Claude Code). It provides copy-paste prompts for common academic writing tasks plus structured workflows for end-to-end paper composition in ML conferences.
How do I install the agent skills?+
Install via npm using 'npx openskills install zechenzhangAGI/AI-research-SKILLs' which places skills in .claude/skills/ where Cursor and Claude Code auto-discover them. For prompts, simply copy-paste the markdown templates directly into your AI assistant chat.
Which AI clients work with these skills?+
The agent skills work with Cursor, Claude Code, and any tool that supports the OpenSkills .claude/skills/ standard. The standalone prompts work with any LLM chat interface including Claude, ChatGPT, Gemini, and DeepSeek.
Do I need API keys or external services?+
For prompts, you only need access to an LLM (Claude, GPT, etc.). For agent skills installation you need Node.js 20.6+ and Git. The skills themselves don't require additional API keys beyond your existing AI assistant credentials.
Is this free to use?+
Yes, the prompts and skills are open-source and free. However, you'll incur costs from your chosen LLM provider (OpenAI, Anthropic, Google) when running the prompts or skills through their APIs.
What models are recommended for these prompts?+
The README recommends Gemini-3-pro/flash for idea interaction and paper writing, Claude-4.5 series for experimental code, and Cursor's Composer for coding tasks. The prompts are tested primarily with these frontier models.
How do I install awesome-ai-research-writing?+
Open the source repository on GitHub and follow its README. awesome-ai-research-writing is a skill — MCP Agents Market links you directly to the official repo.
Is awesome-ai-research-writing free?+
awesome-ai-research-writing is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.