L1B3RT4S

Why we included this project

This repo collects jailbreak prompts organized by vendor, so it works as a reference for anyone who needs to test how stubborn a model's safety filters really are. It is not something you install; you read it, copy the phrasing patterns, and adapt them to the models you are evaluating. That makes it a practical source of candidate inputs for red-teaming suites, and it is handy for comparing how different systems handle the same attack or spotting which prompt structures still get through. Security researchers and AI safety engineers running their own evaluations will get the most out of it by treating it as raw material rather than a finished product.