AGENTS.md and Debugging with AI

I've spent weeks debugging with AI and realized something: the problem was never the AI — it was me giving it half the context.

— 🤖 AGENTS.md works almost like a README for agents — you tell it which commands to run and which conventions to follow, and it stops improvising in your repo.

— ⚡ Errors now come with a prompt that's almost ready to paste into the agent — you skip having to "translate" the stack trace yourself.

— 🔍 Agent Browser lets you see the component tree without manually opening DevTools.

— 🧠 None of this replaces understanding the error — it just speeds up the cycle if you already know what to look for.

AI doesn't think for you, but if you give it good context, it thinks with you. Have you run an AGENTS.md through your repo yet? 👇