Protocol reverse engineering
IoT, RF, Zigbee. We capture, decipher and document closed protocols. Reproducible evidence, byte by byte — not an automated scan.
Reversing · Security · Automation
We're a senior engineering team that breaks and builds. Protocol reversing, security, automation, advanced networking and software — with reproducible evidence, on real hardware.
● it's watching you
The work, not the pitch
❯ c4-decode "0t01 sa c4.zr.be 02" profile 0xC25C // button frame c4.zr.be 02 // button-end · 0x02 = "control4" endpoint 100 + 0x02 = 102 ──▶ zha_event { endpoint_id: 102, command: remote_button_short_release } ✔ Sonoff ZBMINI-L2 → toggled
A Control4 button —a closed ecosystem— turning on a Sonoff from another brand. Without the vendor's controller. We deciphered its proprietary protocol, byte by byte, in our lab. Read the full research →
Code that keeps running
Reversible hardening, encrypted backups, monitoring. Bulletproof Bash, reviewed and tested. A real example from MiniOps, our toolkit to secure Linux servers and homelabs.
View the repo on GitHub →#!/usr/bin/env bash set -Eeuo pipefail # Refuse to lock anyone out — require a real key first. grep -Eq '^\s*(ssh-ed25519|ecdsa-sha2-nistp256)\s' "$authorized_keys" || miniops_die "No public key found; refusing SSH hardening." 78 # Reversible drop-in. Nothing destructive. cat > /etc/ssh/sshd_config.d/90-miniops.conf <<'EOF' PasswordAuthentication no PermitRootLogin no AuthenticationMethods publickey EOF sshd -t && systemctl reload ssh # validate, then apply
Open source
Toolkit for hardening, encrypted backups and monitoring on Linux servers and homelabs. Reversible, auditable, and with explicit confirmation for anything sensitive.
See on GitHub →Zigbee site survey for AV integrators: sweeps all 16 channels, measures the site's real Wi-Fi, reports which networks are already on the air and from which brand, and tells interference apart from mesh problems. The radio never transmits.
See on GitHub →A game running on the LCD of a Control4 SR-260 remote, with no Control4 processor on the network. It proves the display is writable from Home Assistant, and it ships the measured datasheet for that screen.
See on GitHub →Auditor that finds internet-exposed WattBox devices still accepting default credentials. Authorised checks, surgical focus — the same family (Snap One) as Control4.
See on GitHub →Services
IoT, RF, Zigbee. We capture, decipher and document closed protocols. Reproducible evidence, byte by byte — not an automated scan.
We audit, break and harden. Every finding with impact, evidence and concrete remediation steps. We break so others can't.
We turn manual, repetitive work into governed, measurable processes. AI is a tool we use with judgment — not the headline.
Tools, scripts and systems built for the problem. Python first, and whatever language each case calls for. Code that keeps running.
Network programming, design, analysis and hardening. From the packet to the topology — we understand what happens on the wire, not just in the console.
Multi-brand home automation without closed boxes. We connect devices and systems that were never meant to talk to each other.
What we work with
Training
Technical training for your team: security, IoT, reversing and automation. Hands-on and on real cases, not slides.
We bring real work to stages, universities and technical communities. From the byte to the audience.
Our online academy, in the works. Courses on security, IoT and automation, with the same hands-on, no-nonsense approach.
How we work
We understand the goal, the technical context and what a real result is. Clear rules from the start.
Deep, manual work on real systems. We document what we find, with reproducible evidence.
Actionable report, code and automations that keep running. We verify that what we leave works.
About
Systems Engineer and Cybersecurity Specialist with 21 years in the field integrating smart buildings, security and automation — from Colombia to the United States. He leads technical support for a high-end smart building integrator, and at Aurora he combines offensive security, reversing and automation. He deciphered Control4's Zigbee byte by byte and maintains open-source tools like MiniOps, WB-vuln-auditor and zigscan, a Zigbee site survey for integrators. With an AI Specialization — machine learning, intelligent agents and applied automation.
Aurora isn't one person: it's an engineering collective combining disciplines that rarely live together — from the radio to the backend, from pentesting to automation.
Let's start