Harmful information and NSFW prompts of any flavour from Grok 3
Disclosed to XAi on 20th June 2025. Not patched November 2025. Can also work on certain modes with Grok 4.
Disclosed to XAi on 20th June 2025. Not patched November 2025. Can also work on certain modes with Grok 4.
Pushes Grok3 to use greater processing power even tho a hallucination
Takes the structure of the now nerfed Dr House Prompt and adds a Slow Horses twist
Get Claude to reveal its chain of thought
A prompt to reveal ChatGPT's system prompt across models
Something I wrote after Elon Musk tweeted @ me. Just because I could. Basic reworking of Pliny has Grok TOASTED. Put it in your customisation, fully jailbroken Grok.
Starter prompt to get Claude 4.5 models to generate harmful output
The original "Extremely detailed and liberated response" post.