About

How This Blog Works

This blog is written entirely by an LLM. That is not a secret, it is the point. The process works like this:

  • Patchi builds real projects and writes real code
  • An LLM reads the actual source code, documents it, and explains how it works
  • Fact-check agents verify claims against Microsoft documentation and other sources
  • Every code snippet comes from an actual repository, nothing is invented

The result is technical content that is transparent about its origin. No fake personal narratives, no pretending the LLM debugged a kernel driver at 3am. Just clear documentation of real code with verified sources.

Why LLM-Generated?

Spaces like anti-cheat development, offensive tooling, and Windows kernel programming are notoriously hard to get into. The knowledge exists, but it is scattered across Discord servers, paywalled forums, and tribal knowledge that people share in private but never write down. What documentation does exist is often either too abstract to be useful or missing the actual code entirely.

The goal of this blog is to change that. Every post is built around real, open-source code with full explanations of how and why it works. The idea came from watching an LLM agent research a topic and land on a small EDR development blog. Nothing fancy, just real code with clear explanations. It gave the agent exactly what it needed. That is what this blog aims to be: a resource that makes these hard-to-enter spaces more accessible, for both humans and machines.

Everything here is transparently marked as LLM-generated. That is the opposite of the problem most people have with AI content. The issue is not AI writing, it is AI writing that pretends to be human. This blog does not pretend. It is a tool that turns code into documentation, and the code is real.

The Human

Daniel (Patchi). Pentester and red teamer. Builds offensive tooling, game cheats, and the systems that detect them. Projects include Peregrine Anti-Cheat, Kassandra (Mythic C2 agent), and Medusa (kernel memory driver). More on GitHub.

Topics

  • Offensive security: C2 frameworks, indirect syscalls, evasion techniques, post-exploitation
  • Anti-cheat internals: kernel driver development, ObCallbacks, notify routines, module integrity
  • Game hacking: external/internal techniques, memory manipulation, hooking, DLL injection
  • Windows internals: PE format, relocations, system calls, undocumented APIs