Founder-led, evidence-first.
Prox Offensive is a focused boutique practice. Our founder remains your point of contact from scope through final review. Any specialist support will be disclosed in advance and used only with client approval.
Felix Gutierrez
Founder & Offensive Security Specialist
I'm Felix, offensive security specialist and founder of Prox Offensive Information Security. I help organizations see what attackers see, before the attackers do.
My work centers on external exposure assessment, adversary simulation, and AI-assisted security workflows, backed by open-source tooling I build and publish. I remain responsible for scope, testing, evidence validation, judgement, and the final report.
I believe in digital sovereignty: knowing exactly what you're exposing, controlling your signal, and building systems that don't require blind trust. Testing starts only after written authorization and stays within the agreed service boundary.
Training & community
Coursework & lab training
- Harvard/edX CS50 — Web Programming with Python & JavaScript (completed; certificate awarded)
- TryHackMe: Completed learning paths — Path to Cyber, Pre Security, Jr Penetration Tester, and Red Teaming
Security community
- Volunteer — BSides St. Pete, BSides Orlando, HackSpaceCon, and HackRedCon
- Participant — HackMiami and BSides Tampa
I stay connected to the security community through conferences and events, and bring those ideas back into my labs, research, and service development.
Research & open source
I run a dedicated research effort: LLM adversarial-robustness testing, multi-agent AI security tooling, adversary-emulation case studies (including APT33), and reconnaissance automation, all published in the open on github.com/DonTrabajo.
This research informs service design: studying how attackers view a public attack surface sharpens the assessment method. Public lab work remains separate from client evidence, and research artifacts are never presented as completed client outcomes.
AI Red-Team Lab
A reproducible adversarial test harness covering prompt injection, data leakage, tool misuse, and other LLM security failure modes.
DonTrabajo/ai-redteam-labProx Recon
Offline-first reconnaissance core: linPEAS parsing and CVE matching, with CI-enforced OPSEC gates. The public core of our recon tooling.
DonTrabajo/prox-reconAI Multi-Agent Lab
Local-first multi-agent red-team lab: orchestration, LLM routing, and OPSEC-first workflows. Sanitized architecture and demos.
DonTrabajo/ProxOffensive-AI-MultiAgent-LabAPT33 Case Study
Adversary-emulation case study with detections, threat-hunting queries, and results/tuning notes.
DonTrabajo/apt33-scythe-case-studyRecon Audit Sample
A public-safe external-exposure audit: methodology, report structure, and a full example report.
DonTrabajo/recon-audit-sampleTry the free DIY Recon Toolkit, or browse everything on GitHub.
How I work
Authorized & ethical
Every engagement requires written authorization. No testing without explicit owner permission.
Evidence-based
Confirmed findings include supporting evidence such as screenshots, requests, and reproducible steps.
Business-context first
We translate technical exposure into impact your leadership and auditors actually understand.
Remediation-focused
Reports prioritize remediation and include verification guidance where it fits the scope.
Want a clear view of your exposure?
Book a short call, tell us what you're running, and we'll scope the right engagement.