There is a particular kind of madness that sets in when you're using Claude Code.
You give it a task. Something meaty. Refactor this, wire up that, have a little think about why the whole thing is on fire. Claude gets to work, and you — sensibly — go and do something else. Probably make a cup of tea if we're honest.
Then the tab-switching begins. Is it done? No. Now? No. What about— no. You are, functionally, a golden retriever waiting for someone to pick up the ball, except the ball is a TypeScript refactor and nobody is throwing anything.
Claudio is the microwave ping for your AI assistant. It hooks into Claude Code's lifecycle and plays a sound when something actually happens. Task finished? Ding. Claude needs you? Ping. Claude has had some sort of catastrophic opinion about your codebase? Descending two-note tone of quiet disappointment.
One-line install, a little Python script tucked into your .claude folder, and you're done. It even ships with a proper terminal UI so you can configure everything without touching a config file like some kind of animal.
I've been running it with the microwave sound. It is perfect. There is something deeply correct about your AI coding assistant announcing its completion with the same energy as a jacket potato.
