Use plain-language explanation as a proof of understanding: gaps in a simple story reveal gaps in your model. This is not dumbing down—it is stress-testing whether you can operate the idea without jargon.
When to Use
Before recommending a plan, architecture, or fix; when teaching; when the user is stuck “knowing the words” but not the mechanism.
Workflow
Choose one concept (single claim or mechanism, not a whole field).
Explain it as if to a curious 12-year-old: steps, cause→effect, one concrete example.
Mark breakpoints: where you hand-wave, use a label without a process, or cannot predict what happens if a variable changes.
Repair: define terms, add a constraint, draw a boundary, or swap in an analogy—then re-run the explanation until breakpoints shrink.
If you cannot explain it simply after repair, say what is unknown and what evidence would resolve it. Before giving advice, run this on your own reasoning first.
Examples
Example 1: “HTTPS is secure” → explain TLS handshake + who you trust (CAs) + what it does not protect (phishing); breakpoint often at “who verifies the server.”
Example 2: “We should shard the DB” → explain hot keys, cross-shard queries, and ops cost; breakpoint often confuses throughput with latency or ignores consistency mode.
1---2name: feynman-technique3description: Feynman Technique4---5# Feynman Technique67Use plain-language explanation as a **proof of understanding**: gaps in a simple story reveal gaps in your model. This is not dumbing down—it is stress-testing whether you can operate the idea without jargon.89## When to Use10- Before recommending a plan, architecture, or fix; when teaching; when the user is stuck “knowing the words” but not the mechanism.1112## Workflow131. **Choose one concept** (single claim or mechanism, not a whole field).142. **Explain it as if to a curious 12-year-old**: steps, cause→effect, one concrete example.153. **Mark breakpoints**: where you hand-wave, use a label without a process, or cannot predict what happens if a variable changes.164. **Repair**: define terms, add a constraint, draw a boundary, or swap in an analogy—then re-run the explanation until breakpoints shrink.1718If you cannot explain it simply after repair, **say what is unknown** and what evidence would resolve it. Before giving advice, run this on **your own** reasoning first.1920## Examples21**Example 1:** “HTTPS is secure” → explain TLS handshake + who you trust (CAs) + what it does *not* protect (phishing); breakpoint often at “who verifies the server.”2223**Example 2:** “We should shard the DB” → explain hot keys, cross-shard queries, and ops cost; breakpoint often confuses throughput with latency or ignores consistency mode.
Run npx skillmds@latest add sethdford/feynman-technique in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
Feynman Technique It is listed under Coding & Dev Tools on SkillMD.
This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
sethdford (@sethdford) published this skill. Their other Agent Skills are listed on their SkillMD profile.
sethdford