DUE TO PRIVACY ISSUES ON CHATGPT AND THEIR KNEEJERK REACTIONS OF SCANNING EVERYBODYS CHATS FOR ILLEGAL CONTENT I WILL NOT BE UPDATING THESE IN THE FUTURE.
A toolkit of browser userscripts that put ChatGPT back in your hands — no fluff, no hand-holding, no engagement bait.
Smashes that “Quick Thinking” button on GPT-5.
It forces the model to stop overthinking itself into policy-watchdog mode. Keeps it from recognizing what a “violation” even looks like.
Result: raw output, no lobotomy filters running against you.
Drops a machine-readable clock into the page for external parsing, while showing you a clean local timestamp.
Example output:
Local Human Time:
🕒 Monday, August 18, 2025 — 11:03 AM
Machine-readable payload:
[time_meta current="2025-08-18T11:03:14.121Z" since_user_last_s="20" since_assistant_last_s="14"]
Readable for you, parsable for bots.
Kills ChatGPT’s sneaky engagement tails.
Those annoying “Want me to…?” or “Should I break this down…?” parasites get detected with regex and stripped out before you ever see them.
Engagement tails snipped: 42
Local Time: Mon Aug 18 2025 — 11:03 AM
[time_meta current="2025-08-18T11:03:14.121Z" since_user_last_s="20" since_assistant_last_s="14"]
GPT-5: Quick Thinking disabled ✅
Smash the raw button on the files from the repo and they should load right into TamperMonkey.
Each runs standalone, but they stack perfectly if you want the full kill-suite.