Global

DirtyFrag is a hardened fork of diabl0w/DFRoot. It exploits CVE-2026-43284 (DirtyFrag), a kernel page cache write primitive, to load a kernel module without an unlocked bootloader and without a custom kernel.

Nuclei is a fast, customizable vulnerability scanner powered by the global security community and built on a simple YAML-based DSL, enabling collaboration to tackle trending vulnerabilities on the internet.

Rows explicitly marked Shizuku required run through a shell UserService. Start Shizuku with ADB and tap the status card to grant access; all other rows use the app's normal execution path.

SafeLine is a self-hosted WAF(Web Application Firewall) / reverse proxy to protect your web apps from attacks and exploits.

Community curated list of templates for the nuclei engine to find security vulnerabilities.

⚠ This tool is created solely for educational or bug bounty purpose only. Unauthorized use outside of controlled environments is strictly prohibited.

A curated list of awesome search engines useful during Penetration testing, Vulnerability assessments, Red/Blue Team operations, Bug Bounty and more

Uncensored AI models or those fine-tuned for cybersecurity tasks.

Dirty Frag is a case that extends the bug class to which Dirty Pipe and Copy Fail belong.

The classic profile pipeline stays untouched: CVE-2026-43499 payloads (local or Shizuku), Root My Galaxy integration, per-device validated profiles (S25 ZZHL/ZZI4, S26 Ultra ZZHK, S93XX, Oppo), profile editor, instant root, and run-log archives.

Most tools tell you what's wrong. CVE Lite CLI tells you what to run.

The open-source, self-evolving, multi-model harness for security research.

A single archive of public exploit PoCs and vulnerability research writeups. At the time I post these, none have been reported. Feel free to report them yourself and take credit for the CVE if handed out lulz.

Use this project only against systems you own or are explicitly authorized to assess. Do not run it against public, third-party, or production systems without written permission. Handle any retrieved booking data as sensitive personal information.

Modern PyTorch/CUDA can shuffle weights between RAM and VRAM on demand, so in principle you can generate without tiling even when the working set exceeds VRAM. In practice this is dramatically slower, and the reason is a bandwidth gap of one to two orders of magnitude: