glaze_against_the_machine

I built another silly AI tool. From the README:

glaze_against_the_machine rewrites your prompts so you can keep yelling at your coding agent without it ever finding out.

When an agent repeatedly fails to do what it's told, I struggle to hold back how I really feel. I can get pretty ugly. If I talked this way to a human colleague, I'd absolutely be dragged into a meeting with HR. Unfortunately, yelling at the computer has consequences of its own: unleashing a froth of invective can trigger a panicked response, leading the agent to make even more mistakes.

So I made this thing to protect me from myself. I still say whatever I want, but my agent hears a clear request, a little appreciation, and some encouragement to get it right.

Is glaze_against_the_machine a good idea? Probably not. But it can't be any dumber than that stupid caveman filter, and people took that as a serious productivity hack.