Makes Mistakes

Drops the hedge from the ChatGPT, Claude, Gemini, Grok and Google AI disclaimers, in your language. Plus a double-check button.

Every AI chat hedges the same way at the bottom of the screen: "ChatGPT can make mistakes." Can. As if it were a possibility rather than a Tuesday. Makes Mistakes drops the "can". The footer now reads "ChatGPT makes mistakes" 🤡 — same disclaimer, more honest, and gilded with a slow gold shimmer so you actually notice it. Grok ships no disclaimer at all, so the extension writes one for it. It also adds one button above the composer: Improve answer. Click it and your prompt gets a firm reminder appended — "YOU'RE WRONG. DOUBLE-CHECK THE INFORMATION. DON'T LIE TO ME!" — the thing you were going to type anyway, on the third attempt, in caps. The button only shows up once the assistant has actually answered. Works on: • ChatGPT (chatgpt.com, chat.openai.com) • Claude (claude.ai) • Gemini (gemini.google.com) • Grok (grok.com) • Google AI Mode (google.com/search) It speaks your language too — the line is rewritten in whatever language the chat is set to, not just English. No accounts, no tracking, no network requests. Everything happens in your browser, and the only thing it stores is whether you want the button to flash when it appears. Open source: https://github.com/xkelxmc/makes-mistakes