Haydn K.

Haydn K.

Ellsworth, Me · Lead Community Operator · The Cyber Ledger

16 followers

About

Results-oriented Cybersecurity Professional with a focused background in Blue Team operations, including incident monitoring, vulnerability management, and digital forensics. Proven experience leading technical training sessions, managing community-led security initiatives, and developing automation tools for Active Directory auditing. Proactive leader committed to continuous technical research and robust incident response using industry-standard tools and frameworks.

Skills

  • Splunk
  • Wireshark
  • Microsoft Defender for Endpoint
  • Jira
  • ClickUp
  • Digital Forensics (DFIR)
  • OpenVAS
  • GVM
  • Nessus
  • Active Directory Security

27 skills in total

Experience

  • Lead Community Operator

    The Cyber Ledger

    Apr 2026 — Present

    Remote

  • Lead Community Moderator & Admin

    Gladys Ajih’s Cyber Careers

    Jul 2025 — Present

    Remote

  • Cybersecurity Intern

    The Cyber Ledger

    Feb 2026 — Apr 2026

    Remote

Recent posts

It's Getting Cloudy

My posting's been a little light this week, but it's for good reason. I've been diving deep into AWS for my next project, and making great progress.

Life's busy, and the project ended up taking longer than anticipated, but it's officially complete. The infrastructure is built, tested, confirmed working, and I just wrapped up labeling nearly 80 screenshots that map out the entire process.

At this point, all that remains is a clean write-up. The notes in this project series are intended to be detailed, step-by-step guides for anyone to follow along and build their own versions (all while keeping it as free as possible). This is the second part of my AWS series as I prep for the AWS CCP, and the next 5 projects are already mapped out with the intention of touching on the core pillars of AWS.

For now, I won't say what this project is, but the walkthrough is coming soon. In the meantime, what's everyone else working on?

View post

Identity Is the New Perimeter: A Defensive Look Inside Scattered Spider

During my latest cybersecurity research, I focused on analyzing the multi-stage attack chain utilized by modern social engineering syndicate Scattered Spider. Because their tactics bypass traditional endpoint boundaries by attacking the identity layer first, mitigating this risk requires an architecture that treats identity as your primary security perimeter.

The implementation focuses on eliminating legacy authentication fallbacks, securing the help desk pipeline, and ensuring continuous visibility into identity telemetry. Here are the core phases of the defensive framework:

Identity Hardening and Phishing Resistance: Standard push notifications and SMS authentication are highly vulnerable to fatigue campaigns and SIM swapping. The first phase of this architecture involves implementing custom authentication strengths within Okta and Entra ID to restrict corporate access exclusively to hardware-bound, phishing-resistant passkeys and FIDO2 tokens.

Help Desk Structural Controls: To eliminate the risk of human manipulation, the IT service desk must be decoupled from manual MFA resets. By integrating ticket systems directly with identity providers, technicians cannot physically trigger a credential or device reset without an out-of-band verification token being validated and logged, removing social engineering from the equation.

Telemetry Ingestion and Ingress Filtering: Detecting sophisticated identity tracking requires ingesting comprehensive directory system logs into a centralized SIEM. By establishing dynamic network zones, the environment actively drops authentication requests originating from residential proxies and commercial VPNs, stripping attackers of the anonymity they rely on to blend in.

Behavioral Endpoint Tracking: Once an attacker compromises an identity, they live off the land using native utilities. This framework implements strict application allowlisting to block unauthorized remote monitoring and management tools, alongside EDR rules designed to flag memory dumps and volume shadow copy deletions immediately.

The full architectural write-up, including the step-by-step policy configurations and the companion purple team assessment scorecard, is documented in my repository here:

https://github.com/HaydnZK/Research-and-Methodology/tree/main/Scattered_Spider_CTI

View post

Case Study: Engineering a Hardened Cloud Perimeter in AWS

During my cloud security research, I focused on the essential defense-in-depth hygiene required when initializing a greenfield AWS organization. The first hours of a cloud environment's lifecycle are the most vulnerable, and this project establishes a hardened, production-grade foundation for future infrastructure labs.

The implementation focuses on eliminating root exposure, implementing scalable access controls, and ensuring continuous visibility. Here are the core phases of the deployment:

Root Hardening & IAM Identity Center: I activated a centralized identity provider instance to manage cross-account access from a single control plane. The unrestricted root identity was reinforced with physical multi-factor authentication and securely locked away, moving all daily operations to a secure workforce directory.

Role-Based Access Control: To prevent privilege creep, I built out isolated operational groups for network engineering, systems administration, and security auditing. These groups were bound directly to AWS-managed predefined permission sets (such as NetworkAdministrator and SecurityAudit), ensuring human operators are strictly confined to their authorized lanes.

CloudWatch Telemetry Ingestion: I deployed global telemetry enablement rules to ingest CloudTrail management events directly into Amazon CloudWatch Logs across all regions. Choosing Managed Log Groups ensures built-in deletion protection and strict cryptographic source validation, keeping the audit trail safe from adversary tampering.

Fiscal Security Controls: I engineered a proactive billing guardrail using AWS Budgets. Setting a tight monthly cost budget acts as an operational canary in the coal mine, notifying security via automated alerts the moment a resource spike or credential compromise is forecasted to breach the threshold.

By validating the deployment through restricted placeholder accounts, the architecture successfully enforces access denied boundaries across unauthorized lanes, ensuring a true zero-trust baseline.

The full architectural write-up, including the step-by-step group mappings and logging configurations, is documented in my repository here: https://github.com/HaydnZK/AWS/tree/main/Step%20One%3A%20Securing%20the%20Cloud%20Perimeter

View post

BSides Maine: Community, Malware, and Identity Residue

Spent an awesome day volunteering and connecting at the inaugural BSides Maine. Taking part as an auditorium logistics volunteer was a fantastic experience, and the event brought together an incredible mix of technical depth and community focus. Here are the major highlights and takeaways that stood out the most.

Exploding npm Malware Trends

Dipendu Kar and Derian Tenglein gave a great presentation on software supply chain security. The statistics are wild: in 2023, around 5,000 npm packages were removed for being malicious. By 2025, that number skyrocketed to over 500,000 malicious packages out of 4.8 million live packages.

Attackers heavily abuse the single-function dependency nature of npm to inject malicious code via typosquatting, dependency confusion, and maintainer account takeovers.

Many payloads trigger instantly via auto-running pre- or post-install scripts.

While AI tools like ChatGPT and Claude are great at decoding bundled binaries, obfuscated and minified code still present massive static analysis detection gaps.

Key blue team defenses include pinning package versions and utilizing flags like ignore-scripts to prevent auto-running code during installation.

Hunting Identity Residue with ProfileHound

Christopher Haller shared some excellent work on post-compromise escalation and identity security. He introduced ProfileHound, a tool built on the OpenGraph framework that extends classic BloodHound functionality.

Traditional BloodHound uses the HasSession edge, which only logs active, live user logons.

ProfileHound queries the C$ share on domain machines (requiring local admin) to map a new edge called HasUserProfile.

By analyzing the existence and timestamps of user profile directories, defenders and red teamers can visually pinpoint the most heavily used profiles. This allows teams to triage and find high-value identity residue like persistent browser data, long-term session cookies that bypass MFA, DPAPI blobs, and notepad backups.

Mindsets and Hardware Hacking

Jayson E Street gave a phenomenal keynote on physical social engineering. He emphasized that threat actors will exploit anyone trying to be helpful, meaning strict adherence to security rules is vital even when someone claims they urgently need help.

Caught the end of Matt Davis's embedded camera hacking session. Seeing the live terminal process the firmware dump and hit the final confirmation screen was an awesome proof of concept to watch in person.

Proactive Networking is a Game Changer

The vendor kiosks provided a great lesson in the power of building connections. I spent time talking shop with representatives from companies like Deer Brook Consulting, Defendify, Rapid7, IntegSec, TrustedSec, OWASP Maine, and the ISC2 Maine Chapter.

If you are heading to a conference soon, removing the friction from sharing your background makes a massive difference:

LinkedIn QR Wallpaper: Setting my phone background to my LinkedIn QR code made connecting instant.

Portfolio Stickers: I handed out custom stickers for my SleuthHound repository that included a QR code and a quick project description, which made a unique icebreaker.

Mobile Resume: Keeping my resume file instantly accessible on my phone meant I could fill out and submit a complete job application email right from the conference floor when an opportunity came up.

Overall, it was a stellar event that brought a ton of value for both technical insights and community building. I am looking forward to applying these defensive concepts to my upcoming lab projects.

View post

Case Study: Active Directory Control Failures and SIEM Visibility Gaps

During my internship with The Cyber Ledger, I worked through an intense Active Directory incident response project called Silent Forest. For this scenario, I stepped into the role of SOC L3 and Detection Engineering Lead to perform a complete post-mortem of a domain-wide takeover.

The attack chain moved fast, escalating from a password spray to full domain control via a DCSync attack and KRBTGT compromise within six hours. What made this a massive learning experience was that the compromise actually survived initial remediation efforts due to an incomplete KRBTGT cleanup.

My job was to audit the telemetry failures and build the blueprint to fix them. Here are the core areas I focused on:

The Telemetry Failure Split: I classified our visibility gaps into three operational buckets. Telemetry Not Enabled (missing PowerShell script block logging and advanced Azure AD auditing), Telemetry Present but Not Correlated (unlinked log anomalies), and Alerts Generated but Not Operationalized (ignored identity alerts).

Endpoint Control Failures: I analyzed how the attacker easily disabled Microsoft Defender real-time monitoring via PowerShell, tracing the issue back to weak GPO privilege governance, lack of change monitoring on security policy objects, and missing tenant-level tamper protection.

The Persistence Flaw: The reason the attack survived remediation came down to a single rotation of the KRBTGT account. Because the initial team didn't perform a double rotation, previously forged Golden Tickets remained valid within their lifetime window, giving the attacker a quiet backdoor right back into the environment.

To bridge the gap between technical detection and business risk, I authored a comprehensive 30-60-90 Day Security Hardening Plan mapped directly to ISO 27001 Annex A controls. The plan prioritizes immediate domain controller log ingestion and DCSync detection rules in the first 30 days, moves into restricting replication permissions and expanding PowerShell transcription by day 60, and finishes with purple team validation exercises by day 90.

I also built a SIEM Improvement Blueprint to shift our monitoring away from isolated alerts and move toward behavior-based correlation rules that focus on high-risk, identity-centric attack paths.

The project was a fantastic exercise in looking beyond the malware to fix the underlying identity and engineering gaps that let an adversary succeed.

You can find the full breakdown, the visibility matrix, and the exact 30-60-90 day hardening roadmap in my repository here: https://github.com/HaydnZK/TCL-Internship/tree/main/Challenges/Silent_Forest

View post

View Haydn K.'s profile on Cyberr