I turned Andrej Karpathy's tips on understanding LLM output into an Open Source Agent Skill : [make-it-click](https://github.com/ajithraghavan/make-it-click) 🧠⚡
Inspired by his X post : as Models do more of the work, we spend more time understanding it. So the format of an answer matters
The Skill picks the simplest format that makes an idea click, and moves up only when needed :
Bonus : an oversight mode for reviewing Agent work (what changed -> why -> risk -> what to check) MIT licensed. Works with Claude.ai and Claude Code
🔗 https://github.com/ajithraghavan/make-it-click Feel free to use and ⭐️ the Repository
Inspired by @karpathy: [https://x.com/karpathy/status/2105819303471976479] Thanks for Reading!
source & further reading
dev.to — original article
AI Turned an Assumption into a Rule, and Tomorrow Nobody Will Know Why.
Is the model actually getting dumber, or are we just reading tea leaves from single samples?
Iam 12 .I Accidentally Built a Secure JS Sandbox While Patching a Bug. Here is How KODA Runs AI Code.