About

I’m Daniele Linguaglossa, better known as dzonerzy - an offensive security researcher and developer.

I spend my time taking systems apart until they confess: source audit, root cause, working exploit, write-up. Kernels, userspace daemons, Windows internals, embedded devices, and lately machine learning models - if it parses untrusted input, it’s in scope. Some of that work ends up on this blog; the rest ends up in coordinated disclosure.

Projects

Here is a list of the projects that I made in my spare time, some of them are still in development, but I’m looking forward to finish them all.

libdzonerzy.so – libdzonerzy.so is a shared object that I wrote in C, it’s the library that I use to power this blog.

Series

Collections of articles grouped under one hat.

PROMPT & PWN – Can AI find real, impactful 0days? An honest scoreboard of the effective capabilities - and the limits - of AI in vulnerability research, one exploited bug at a time (2 articles)

Articles

Below you can find a list of my articles, I’m writing about offensive security and programming, and also about my projects.

2026

PhantomLayout - Ghost Layouts, Real SYSTEM – How a poisoned keyboard-layout descriptor in the shared window-station IME cache became a handle-less cross-process injection primitive (no OpenProcess, no WriteProcessMemory, no remote threads, no hooks), and how a message-free foreground-layout chain rode it from a medium-integrity user to SYSTEM inside consent.exe, before the UAC prompt was even answered #windows #ime #injection #privesc
HeartAttACK: An SCTP HEARTBEAT ACK Control-Queue Use-After-Free – How a parked SCTP heartbeat reply kept a raw pointer to a transport that ASCONF had already freed, and how I rode that 14-year-old lifetime bug to a deterministic, race-free local root on the Ubuntu 7.0.0-28-generic kernel - whole-slab cross-cache recycling, a verification-tag overlap oracle, and a rewritten f_mode in struct file. No ROP, no shellcode, KASLR untouched #kernel #linux #sctp #uaf #privesc
Neural Backdoors - When Your AI Has a Secret Agenda – A weekend journey into neural network security, poisoned datasets, and why detecting backdoors is harder than you think #ai #neural-networks #backdoors

2023

From zero to botnet – GL.iNet going wild (RCE + Botnet) #iot #router #rce #botnet

Advisories

Below you can find a list of my CVEs.

CVE-2023-46453 – GLiNet Router Authentication Bypass + RCE

FAQ

Questions I get asked a lot. The short version of most answers: dzonerzy@gmail.com.

Q: How can I contact you?

A: Email me at dzonerzy@gmail.com. Research, bugs, collaboration, responsible disclosure, or just to talk shop - all welcome.
Q: What do you do for a living?

A: Offensive security research. Who signs the paycheck stays off the record - opsec is a lifestyle.
Q: Where are the PoCs?

A: Exploit code gets published once coordinated disclosure runs its course, not before. If something I wrote about affects you, the remediation section of the post is the place to start.
Q: Can I republish or translate your articles?

A: Ask first, link back, and we’ll most likely be fine.