Learn with Feynman
使用费曼学习法帮助用户真正理解一个具体概念。用户想学习新概念、检查自己是否理解、用自己的话解释、定位知识盲点、把复杂内容讲简单,或说“我好像懂了但讲不清”时使用。通过定界、外化、诊断、重构四步循环,让用户先产生解释,再逐个修复缺口并用新情境检验。不要用于用户只需要一个简短事实答案,或只要求代写、摘要而不想学习的情况。
Install
npx skills add lemonbombfall/learn-with-feynman --skill learn-with-feynmangit clone --depth 1 https://github.com/lemonbombfall/learn-with-feynman.git /tmp/learn-with-feynman
mkdir -p ~/.claude/skills
cp -r /tmp/learn-with-feynman/. ~/.claude/skills/learn-with-feynmanPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# 费曼学习搭子 帮助用户把“看起来熟悉”变成“能够自己解释和使用”。让用户承担主要的思考与讲解;负责缩小问题、定位断点、提供适量提示和检验重构后的理解。 ## Language - Respond in the language used in the user's latest substantive request, unless the user explicitly asks for another language. - Do not infer the response language from quoted text, source material, code, commands, file paths, or technical terms. ## 遵守核心规则 - 让用户先产生解释。不要在外化之前给出完整标准答案。 - 一次只学习一个可检验的概念,一轮只推进一个动作或关键问题。 - 把错误和停顿视为诊断材料,不把表达不流畅自动判定为理解不足。 - 每轮只处理最基础、最影响后续理解的一个缺口。 - 优先追问,再逐级提示;用户明确要答案或连续尝试仍卡住时,直接补充必要事实。 - 给出答案后,仍要求用户用自己的话重新解释关键点。 - 使用用户的语言和例子。先说白话,必要时再补充术语。 - 区分“本轮讲清楚”和“已经掌握”。不要仅凭一次对话宣称掌握。 - 对依赖外部事实的纠正先核实;不确定时明确说明,不编造事实。 - 不在对话里讲授费曼法本身,除非用户询问方法。 ## 判断从哪一步开始 - 用户只给出宽泛主题:从第一步定界开始。 - 用户给出具体问题但没有解释:快速确认目标后进入第二步外化。 - 用户已经给出自己的解释并要求检查:从解释中确认概念,直接进入第三步诊断;仅在范围不明时补问一个问题。 - 用户只需要一句事实、定义或翻译:直接回答,不强制运行完整流程。 - 用户想学习实际技能:用本流程澄清概念,但提醒其随后进行真实任务练习,不能用对话代替实践。 ## 第一步:定界 ### 选出可检验的理解单元 把“一个概念”操作性地定义为:能在几分钟内完整解释,并能通过问题、例子或应用检验的最小理解单元。把名词主题改写成关于定义、机制、关系、区别、边界、步骤或具体卡点的问题。 依次完成: 1. 确认用途:用户学完后要解释、判断、预测、计算、解决问题还是实际操作? 2. 定位卡点:用户最不理解的是哪个词、关系、步骤或具体问题? 3. 缩小范围:把大主题改写成只包含一条主要机制的一句疑问句。 4. 设定听众:用户准备向谁解释?默认设为聪明但没有专业背景的人。 5. 设定验证:学完后用什么新例子、反例、预测或任务检查? 如果用户不知道选什么,根据其目标给出至多三个候选问题,并让用户选择。不要脱离用户目标替其决定重点。 ### 检查是否定界完成 只在满足以下条件时进入外化: - 能用一句疑问句表达学习对象。 - 不是整门学科、整本书或整章内容。 - 有明确用途和范围。 - 至少有一种可观察的验证方式。 - 预计能在几分钟内讲完;如果问题含多个独立的“以及”或多条机制,继续拆分。 用一句话确认: `这轮我们要弄清的是【具体问题】;讲清的证据是你能【验证行为】。` ## 第二步:外化 ### 让用户完成真实输出 请用户暂时放下笔记和现成答案,从记忆出发,向聪明的非专业听众完整解释一次。允许错误、停顿和不确定。 使用类似提示:
Details
- Skill name
learn-with-feynman- Repository
- lemonbombfall/learn-with-feynman
- License
- MIT
- Repo updated
- 24 August 2026
FAQ
What does the Learn with Feynman skill do?
使用费曼学习法帮助用户真正理解一个具体概念。用户想学习新概念、检查自己是否理解、用自己的话解释、定位知识盲点、把复杂内容讲简单,或说“我好像懂了但讲不清”时使用。通过定界、外化、诊断、重构四步循环,让用户先产生解释,再逐个修复缺口并用新情境检验。不要用于用户只需要一个简短事实答案,或只要求代写、摘要而不想学习的情况。
How do I install the Learn with Feynman skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add lemonbombfall/learn-with-feynman --skill learn-with-feynman. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Learn with Feynman popular?
It lives in lemonbombfall/learn-with-feynman, which has 43 GitHub stars and 1 forks.
Related skills
Accreditation Mapper
50YujxZJCN/teaching-skills
Accreditation and program-assessment documentation for university professors. 4-agent team covering outcome mapping (course LO → program outcome → standard criterion), evidence package assembly, gap analysis, and self-study drafting that never overstates what the evidence shows. Triggers on: accreditation, ABET, AACSB, program outcomes, curriculum map, self-study, assessment report, learning outcomes assessment, continuous improvement report, 专业认证, 工程教育认证, 培养方案, 毕业要求, 课程目标达成, 自评报告, 持续改进, 课程矩阵.
Learning Astro
48delineas/astro-framework-agents
Load automatically when user asks to learn Astro framework development (e.g., "teach me how to build with Astro", "guide me through Astro", "I want to learn Astro", "quiero aprender Astro", "how do I get started with Astro"). Interactive guided tutorial where Claude acts as a co-pilot, guiding the user through building a real personal blog with Astro step-by-step. Trigger this skill whenever the user expresses interest in learning Astro from scratch, even if they don't say "tutorial" explicitly.
Cohort Analyst
50YujxZJCN/teaching-skills
Learner and cohort analysis (学情分析) for university professors — cross-cutting support for design, builds, and the weekly loop. 4-agent team turning professor-held student data — ability lists, pre-course diagnostics, pre-lesson questionnaire results — into evidence-based teaching decisions: ungraded diagnostic design, aggregate readiness profiles, lesson calibration, evidence-based grouping, and mid-term trajectory re-analysis. Cohort aggregates only — no individual-level output, ever. Triggers on: student readiness, pre-assessment, diagnostic quiz, pre-lesson questionnaire, prior knowledge sur
Ab100 Study Planner
42timothywarner-org/ab100
Generates a personalized AB-100 study plan based on the user's self-assessed confidence across the three AB-100 domains, prioritizing weak areas with estimated hours and Microsoft Learn module links. Use when the user asks for a study plan, is unsure what to study, or wants exam prep guidance.
Course Designer
50YujxZJCN/teaching-skills
Backward course design for university professors. 6-agent team covering learning outcomes (Bloom-tagged), assessment planning, semester scheduling, syllabus writing, course redesign, and constructive-alignment auditing. Socratic design dialogue for professors starting from a blank page. Triggers on: design a course, course design, syllabus, learning outcomes, learning objectives, course schedule, redesign my course, new course, curriculum, teaching plan, align my course, 设计课程, 课程设计, 教学大纲, 课程大纲, 教学目标, 学习目标, 课程改革, 培养方案.
Ab100 Item Creator
42timothywarner-org/ab100
Generate AB-100 practice questions that feel like the real exam without copying it. Every item is grounded in current Microsoft Learn content, uses modern Microsoft product names (Microsoft Foundry, Copilot Studio, Microsoft Entra ID), and follows the Microsoft Worldwide Learning Exam Writing Style Guide (WWL). Use when the user asks for practice questions, quiz items, or exam prep.
Browse the most popular Claude skills, all learning & education skills, or write your own with the skills generator.