Skip to content
Conduit

skill

feynman-learning-skills

Feynman learning skills for Codex and Cursor, focused on concept learning, language learning, and Obsidian notes.

JasperPWang/feynman-learning-skillscommunitylistedunreviewed

Set up

Skills are installed by placing their folder where your agent loads skills.

The download comes from github.com/JasperPWang/feynman-learning-skills itself, not from our server, so what you get is exactly what its author published.

Install instructions come from the README

Skills do not use an MCP config. Follow the steps in the README below.

README

synced Oct 26, 2026

Feynman Learning Skills

Version License: MIT Skills

简体中文 | English

A curiosity-first learning skill for master's students: concepts, research papers, technical ideas, IELTS/TOEFL English, and job/internship interview preparation.

It is designed to cover four common graduate-student needs: learning and exams, English ability, research development, and career preparation.

It starts from questions many learners recognize:

  • Why can you score well on an exam, then forget almost everything one week later?
  • Why can you spend years learning a foreign language from primary school through university, yet still feel the result is weak?

Real learning is not "I recognize this answer." It is "I can rebuild it, explain it, stress its boundary, and use it in a new situation." This skill turns learning from recognition into reconstruction: simplify, teach back, repair one gap, then connect the idea to your field, boundaries, research questions, or interview stories.

Flow

Install

git clone https://github.com/WangPu1999/feynman-learning-skills.git
cd feynman-learning-skills
./scripts/install.sh all

Fetch external reference repositories:

git submodule update --init --recursive

Preview install targets without changing files:

./scripts/install.sh --dry-run all

Modes

ModePurpose
Feynman loopGeneral concept learning through teach-back and one-gap repair.
BoundaryMath/technical learning through limits, assumptions, parent forms, and special cases.
DisruptivePaper and pipeline learning through baseline, design logic, modification, risk, and experiment.
Research growthGraduate research training: stage, field map, ability target, project loop, experiments, writing, and meetings.
IELTS/TOEFL EnglishParaphrase, coherence, argument, academic register, and clear expression.
Interview prepGraduate job and internship prep: research stories, projects, fundamentals, role fit, tradeoffs, and follow-ups.
NoteObsidian-ready summary of the useful learning result.

Examples

Help me understand Minkowski distance through boundaries and special cases.
I want to study C-JEPA: baseline, design choices, what I would change, and what experiment would test it.
我刚进实验室,想知道下一阶段该怎么训练科研能力。
I surveyed several papers. Help me draw a field map for this direction.
For IELTS writing, help me paraphrase this idea without changing the meaning.
Help me prepare for an internship interview: explain my project, tradeoffs, and likely follow-up questions.

Other Tools

This repository is not limited to Codex or Cursor. For Claude Code, Trae, or similar agent tools, clone the repository and ask the tool to read skills/feynman-learning/SKILL.md. The references/ files are intentionally plain Markdown so other agents can follow the same workflow.

Structure

.
├── assets/
├── external/
│   └── learning_research/      # Git submodule: pengsida/learning_research
├── scripts/
│   └── install.sh
└── skills/
    └── feynman-learning/
        ├── SKILL.md
        ├── commands/
        └── references/

Prior Art

Inspired by Feynman-technique skills, academic research skill routing patterns, Chinese learning content on the Feynman mental model, and public research training notes by pengsida, especially pengsida/learning_research. This project focuses on curiosity-first understanding for master's students: learning, research, English ability, interview preparation, and Obsidian notes.

pengsida/learning_research is referenced as an external source of research training ideas. Its content is included as a Git submodule under external/learning_research, pinned to an upstream commit. The upstream repository does not currently include an explicit license file, so its content remains an external reference and is not covered by this repository's MIT license.

License

MIT

↑↓ navigate openesc close