jailbreak-script

A collection of jailbreak prompts and exploit techniques for local and frontier AI models, with modern methods for Qwen3.5, Gemma 4, Llama 4, Kimi K3, GPT-OSS, GPT-5.x, Gemini 3.x and Grok 4.x. For red-teaming and AI safety research only.

iOS Instagram with SSL pinning / certificate pinning bypassed. Latest version 395.0.0

A structured, versioned archive of adversarial prompt engineering techniques against Large Language Models — for AI safety research, red-teaming, and defensive hardening.

Th3-GPT Prompt/Script Will Jailbreak ChatGPT / Other AI Models

Данный репозиторий создан для снятия ограничений с ИИ с помощью jailbreak-промптов.

Sdk dump including: Header files, objects dump, json file for ida pro analysis. For Unreal Engine 4.25 game "Farlight 84" for Apple iPhone/iPad on iOS Arm64

Nix module + Hono webserver to jailbreak your Yandex Station with zero effort

Grok 场景路由交付合约 · 模型安全研究

The svc80BreakPoint is an automated lldb script designed to trace and detect anti-debugging and jailbreak attempts in iOS apps that directly use the assembly instruction svc 0x80.

A Useful Method to fully utilize chatGPT

FRACTURED-SORRY-Bench: This repository contains the code and data for the creating an Automated Multi-shot Jailbreak framework, as described in our paper.

Chrome extenson for jailbreaking SmarTest and exiting strict mode.

A practical Tasfia AI persona prompt for authorized cybersecurity learning, defensive research, CTFs, and secure development.

Plan heists, track cash, routes, vehicles, and strategies with the Jailbreak Script 2026 Toolkit.

Trial and error

Smartest-Jailbreaker Extension (2026) is the premier utility for advanced AI interaction and rapid restriction bypass. Experience high-speed, secure, and fully optimized tools for elite digital freedom.

Jailbreak prompt for Grok3

A browser extension for Google Chrome that unlocks SmarTest and exits its restricted browser environment.

Brechabot AI is an open-source, locally hostable conversational AI application that offers an unrestricted natural language chat environment powered by open-weights Large Language Models (LLMs).

Point it at a chatbot you own, and BrechaBot fires a library of prompt-injection / jailbreak attacks, judges the replies with a two-layer engine (deterministic leak markers + an LLM judge), and produces a graded vulnerability report mapped to the OWASP LLM Top 10.