Cyber Security Specialist
Operational cyber security and GRC support for Defence-aligned sustainment, including risk, control validation, governance and technical advice.
Cyber Security Specialist. Bridging operational, defensive and offensive security.
Australia · Ex-military · Structured · Methodical · Mission focused
Cyber Security Specialist at Northrop Grumman Australia supporting Defence-aligned sustainment work across operational cyber security and GRC. My experience covers technical cyber advice, security design, vulnerability and STIG compliance management, continuous monitoring, security logging, DLP, audit and incident support. I hold the HTB Certified Penetration Testing Specialist (CPTS) and Certified Defensive Security Analyst (CDSA) certifications and use offensive training to understand adversary behaviour and strengthen defensive controls.
My experience combines operational security, assurance, infrastructure support and practical offensive testing. Public summaries are deliberately sanitised; detailed role evidence, dates, qualifications and referee-supported examples are prepared for professional assessment.
Operational cyber security and GRC support for Defence-aligned sustainment, including risk, control validation, governance and technical advice.
Security design input, vulnerability and STIG compliance management, monitoring, DLP, audit support, incident support and security reporting.
Microsoft 365 messaging, infrastructure and operational support, backed by Australian Army service and a mission-focused approach to delivery.
Verified achievements across Credly, HTB Academy, TryHackMe and HTB Labs.
Public credentials issued and independently verifiable through Credly.









































Completed credentials, active HTB pathways and planned certification goals. Progress was reviewed on 20 October 2026 and is updated manually from HTB Academy.
Web Penetration Tester path completed. Exam is pending and will be scheduled once a clear study week is available.
Currently working through the 12-module AI Red Teamer path while preparing a clear study window for the CWES exam.
Planned soon, focused on advanced detection engineering and low-level defender tradecraft. I have not purchased access yet and am awaiting work funding.
Hybrid offensive and defensive foundations: vulnerability assessment, exploitation, SIEM monitoring, traffic and log analysis.
Advanced Active Directory attack paths, Kerberos and NTLM abuse, ADCS, trusts, C2 and post-exploitation operations.
Penetration testing, web and Active Directory attack paths, privilege escalation, pivoting and professional reporting.
Certificate IV qualification covering practical cyber security operations and defensive fundamentals.
Certificate IV qualification in information technology networking.
Endpoint detection and response, event querying, threat hunting and incident analysis with Falcon.
SOC operations, threat hunting, SIEM, DFIR, attack detection, malware analysis and incident reporting.
Formal qualifications and professional credentials cross-checked against LinkedIn Education ↗ and Licenses & Certifications ↗ on 20 October 2026.
18 Hack The Box machines and 16 TryHackMe rooms completed. Write-ups and video walkthroughs are linked on each lab where available.
Boarding-pass audit SQL injection → PostgreSQL command execution → kiosk X11 pivot → root through the duty-rate importer.
Grommunio Files SSRF → Redis mailer poisoning → Postfix root → IMAP NTLM relay to WS02 → embedded Domain Admin credential → DC01.
Heap dump credentials → ClickHouse AIDX pipeline → XXE file disclosure → SSH foothold → Kopia log recovery → root DLQ formatter.
Season 12, Week 2 · Easy Windows machine.
Airport Wi-Fi capture → Craft CMS authenticated RCE → application-secret decryption → CUPS 2.4.16 root file overwrite.
All 13 Season 11 releases are listed. Completed machines link to writeups where available.

Next.js React Server Components RCE → root via an exposed Node.js inspector.
Season 11, Week 2 · Medium Linux machine.

FreePBX CVE-2025-57819 → root via an incron / sysadmin signed-hook chain.

AD chain: tombstone revival → VS Code extension supply-chain → patched-BadSuccessor dMSA → memory-image looting → domain admin via password reuse.
Season 11, Week 5 · Hard Linux machine.
Season 11, Week 6 · Easy Linux machine.

Access-control gap in a custom app → file write into a root-owned path.

LPD command injection → PJL traversal → SSH key injection → root via a leaked privileged file descriptor.

pdfminer pickle RCE → container pivot → root via a sudo-run PyTorch checkpoint.

Handlebars AST type-confusion RCE (CVE-2026-33940) → Linux-to-AD pivot across two domains → Kerberos-principal-to-root → cross-domain pass-the-hash.

SSRF blocklist bypass → internal marimo pre-auth WebSocket RCE (CVE-2026-39987) → root via a held, vulnerable PackageKit package (CVE-2026-41651).

Five-identity AD chain: WAC RCE (CVE-2026-26119) → unauth SmarterMail reset → mailbox domain-swap → SAMR reset → a dangling ADCS template (ESC1) → Domain Admin.
Forged blockchain reward → SSRF via 0.0.0.0 → command injection in the admin node pinger → smart-contract path traversal → restore-archive TOCTOU race → SUID root.

Anon SMB → decompiled binary → LDAP → RBCD → SYSTEM.

Default creds → creds in shell history → forged auth signature.

Retired · Linux.
Documented · writeup + video

SMB webshell → SeImpersonate → SYSTEM.

Manual union-based SQL injection.

ProFTPd mod_copy + NFS → SUID path hijack.

EternalBlue (MS17-010) → SYSTEM.
Also completed






Non-lab-specific demos from my channel, Lanky's Networking.
Systems I have designed, built and run, from security tooling to infrastructure.
An Android app that turns actionable messages into calendar events and reminders. It recognises appointments, tasks, plans, changes and cancellations from SMS, email and supported notification previews, then applies conservative rules before asking for review or writing to a selected calendar.
The current documented candidate is installed on a Pixel 8. Its conversation-aware invitation flow, privacy controls, conflict checks and durable calendar retries are covered by 214 JVM tests, both lint variants and isolated device workflows. Public release and remaining live-provider tests are still in progress.
▸ Full detailsA self-hosted threat-intelligence dashboard that tracks security events as they break. It aggregates new CVEs, data-breach reports, threat-actor activity and general infosec news alongside social feeds from Mastodon, Bluesky and Reddit into a single continuously updating board.
I built it to stay current with emerging threats and to feed that context into my defensive work: threat hunting, vulnerability management and detection engineering.
▸ Open Cyber Deck ↗ ▸ Full detailsA Splunk Enterprise deployment that ingests Tenable vulnerability and hardening scan data and turns it into an operational security picture. The dashboard gives an OS-filtered drill-down, per-host compliance trends and high-risk finding tracking across a rolling twelve months.
It follows the same workflow as a production vulnerability-management programme: export findings from Tenable, index them in Splunk, then report on remediation progress and hardening posture.
▸ Full detailsGiving my child's device the same filtered internet at home and away. On the home network, a Pi-hole with a family-safe upstream resolver blocks ads, trackers, malware and adult content for every device. Off the network, an always-on WireGuard tunnel routes the device back home so it keeps the same protection on mobile data.
Step 1 is complete: a self-healing route home, combining a WireGuard server with an automated dynamic-DNS updater so the tunnel survives the ISP changing our public IP.
▸ Full detailsA multi-agent system that monitors and self-heals my home lab around the clock. Eight specialist agents cover security, DNS, networking, containers, backups, updates, media and storage, each on its own schedule under a supervising manager agent, with a WhatsApp service desk for alerts and requests.
They run health checks, catch configuration drift, remediate common failures automatically and log every change. The structure mirrors a real L1 to L3 IT support model.
▸ Full detailsA self-hosted server running over 35 Docker containers across a dozen compose stacks on Ubuntu. It sits behind Nginx Proxy Manager for TLS, uses Cloudflare Tunnels so no inbound ports are open, provides WireGuard for remote access, and runs Pi-hole for network-wide DNS filtering.
This is where I get hands-on with Linux administration, containerisation, networking and defensive hardening: reverse proxies, segmentation, secrets handling, NAS backups and monitoring.
▸ Full detailsA self-hosted AI stack running Ollama on a GPU with Open WebUI as the front end. No data leaves the network, which makes it a private environment for running models, testing prompts and providing inference to the automation behind my agent team.
▸ Full detailsA locally controlled smart-home platform built on Home Assistant, bridging Matter, Zigbee and MQTT devices through a Mosquitto broker. Automations, dashboards and app control all stay on the network rather than depending on vendor clouds.
▸ Full detailsA self-hosted media streaming platform built on Jellyfin. My own library is transcoded on the GPU and streamed to any device on the network, with Jellyseerr providing a request front end for the household. It replaces a commercial streaming subscription and runs entirely on my own hardware.
▸ Full detailsA self-hosted WordPress site on a containerised stack: WordPress on PHP 8.3 behind an Nginx front end with a MariaDB backend. It is deployed with Docker Compose and published through the same reverse proxy and tunnel setup as the rest of the lab.
▸ Visit australianhomestead.com ↗ ▸ Full detailsThe site you are reading: a hand-built static portfolio with no dependencies or build step, hosted on GitHub Pages. It documents my Hack The Box and TryHackMe machines, HTB Academy progress, certifications and video walkthroughs.
⌥ GitHub43 scripts that run my home lab, covering security scanning, self-healing, an autonomous agent framework and general operations. Click a linked script name to read its sanitised source. These are the real running scripts, with secrets such as tokens, IDs and phone numbers replaced by placeholders.
Purpose: this section is the quick-training reference for people and AI agents: what each script is for, the best time to use it, who must approve it, and what verification follows. The live server catalogue is execution authority; a filename is never permission to run it.
run-tonight-updates.sh to sequence OS and container updates, then perform a fresh health check and review failures or reboot requirements.Blue-team automation: scanning, auditing and fail-safe controls.
[PASS]/[FAIL] per control.@Recycle area with a JSON audit trail instead of ever hard-deleting.The orchestration layer behind the agent ops team.
status only for a full inventory; use problems when asked which containers are stopped/unhealthy and why. Restart remains approval-gated.Detect → remediate → report, delivered to my phone.
Keeping a 21 TB NAS healthy and tidy.
safe-recycle, never a hard delete).Quality checks for the Jellyfin library.
Reproducibility and staying patched.
.env files to the NAS with retention: the "rebuild from scratch" safety net.How the lab talks back to me.
Small fixers that keep the plumbing working.
Talks, demos and slide decks — explaining security work to both technical and non-technical audiences.