A Claude skill that explains any technical topic using a simple short story, so even a 5-year-old can understand.
Curious Kid works with Claude. Ask about programming, ML, networking, finance, medicine, or science, and it answers with a tiny story first, then connects it back to the real concept. You get the shape of the idea in under a minute, so the jargon has something to attach to later.
You: /curious-kid explain what a cache is.
Claude:
The Story Mia sells lemonade. Her jug of lemonade is far away, in the kitchen. Every time someone wants a cup, she runs to the kitchen. It is slow.
So Mia fills a small cup and keeps it on her table. Now when someone wants lemonade, she hands over the cup right away. So fast!
In real life kitchen jug = slow storage cup on the table = cache handing it over right away = fast access
Where it's not perfect A real cache is small, so it can only hold a few things at a time.
The simple truth A cache is a small, fast place that keeps things you use often, so you don't have to fetch them from far away each time.
Want to go deeper? I can explain it again with real terms like cache hits, misses, and eviction.
- Story first. Intuition comes before definitions.
- One idea at a time. Each story covers only the core concept, not the whole topic.
- Honest about its limits. Every answer includes a "where it's not perfect" line, so the analogy doesn't leave you with wrong ideas.
- A path forward. Ask to go deeper and it re-explains with real terminology, one step at a time.
- Zip the
curious-kidfolder. - Go to Settings > Capabilities > Skills and upload the zip.
- Make sure the skill is toggled on.
Copy the folder into your skills directory:
# Personal (all projects)
cp -r curious-kid ~/.claude/skills/
# Or project-specific
cp -r curious-kid .claude/skills/Menu names and paths can change between versions, so check the current Claude documentation if a step doesn't match.
Just ask. Any of these will trigger it:
Curious Kid, explain how DNS worksELI5: what is gradient descent?Explain compound interest with a storyI don't get Docker. Make it simple.
Then follow up with:
go deeperfor the same topic with real terminologyanother examplefor a different story on the same idea
The skill's behavior is defined in SKILL.md. The key rules:
| Rule | Limit |
|---|---|
| Ideas per story | 1 |
| Characters | 2 or fewer |
| Story length | 50-100 words |
| Sentence length | Under 12 words |
| Story parts mapped to real concepts | 3 or fewer |
| Plot | One problem, one fix |
Before answering, the skill runs a self-check and rewrites the story if it breaks any of these limits.
Edit SKILL.md to fit your taste:
- Simpler or richer stories: change the word limit in the Simplicity Rules.
- Different tone: edit the Tone section (more playful, more formal, and so on).
- Domain-specific settings: add guidance such as "for finance topics, use stories about sharing snacks or trading toys."
- Different output sections: change the Output format block.
- Analogies simplify. Always treat the story as a starting point, not the full picture.
- Very broad topics (for example, "how the internet works") get one core idea per story, so expect to ask follow-ups.
- Not meant for precise technical specs, code review, or debugging.
curious-kid/
├── SKILL.md # The skill itself
├── README.md # This file
├── LICENSE # MIT license
└── assets/
└── logo.png
MIT. See LICENSE for details.

