Supervisor Skills

Supervisor Skills: Supervisor-Skills is an open-source library of structured AI prompts and guidelines designed to assist researchers in navigating the full academic lifecycle.

Supervisor-Skills consists of a handbook and a set of executable skills designed for researchers. It translates academic experience from data science and AI conferences into structured prompts that guide large language models like Claude, GPT-4, and Kimi. The project addresses research tasks ranging from idea evaluation and literature surveying to paper writing, formatting, and rebuttal preparation. Users access these skills by importing specific prompts into their preferred AI assistants, such as Claude Code, Cursor, or Codex. This system helps bridge the gap between theoretical knowledge and practical execution for graduate students and researchers.

Supervisor Skills at a glance

Best for: Researcher, Student
Runs on: Web, CLI
Pricing: Free (Open source under CC BY-NC-SA 4.0 license)
Released: April 2026
Category: AI research tool
Website: github.com/HKUSTDial/Supervisor-Skills

Key features

Idea evaluation based on a multi-dimensional framework
Evidence-gated paper drafting and literature verification
Language polishing for academic submissions
Automatic reconstruction of diagrams into editable Draw.io files
Rebuttal guidance based on reviewer concern analysis
Step-by-step logic checking for technical and benchmark papers

Use it for

Drafting six-paragraph introduction sections for research papers
Evaluating initial research ideas against objective criteria
Reconstructing academic diagrams into editable formats
Checking manuscripts against submission guidelines

Alternatives to Supervisor Skills

AI Research Skills — 98 expert skills for training, evaluating and writing up ML research
Scientific Agent Skills — Provides 177 validated skills and 100+ scientific databases for research agents.
anti defensive writing Skill — Stops self-censorship and frames academic papers around core strengths.
Academic Research Skills — Provides Codex-native academic research workflows and automated routing.
OpenClaw Medical Skills — Adds 869 clinical, genomic, and drug discovery skills to OpenClaw and NanoClaw agents.
ResearchRabbit — Access over 310 million academic papers and map research fields visually.
ARIS — Cross-model review loops verify research steps without framework lock-in.
DocTree — Generates source-grounded learning articles adapted to your expertise and time
Kelip Paper Reading — It breaks down papers across six detailed stages with local source code and image extraction.
Paper Xray — Recovers hidden failure modes, author bets, and concrete symbol examples.
phil-OKF: any topic, one reading guide — Generates minimal research-paper styled study pages with verified source citations.
PaperLeet — Practice implementing 89 Python exercises derived directly from AI research papers.

All Supervisor Skills alternatives