CVE Activity
Loading threat feed…
── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ── ──

Thassanai McCabe

Cyber Threat Intelligence Researcher — Ireland
ReliaQuest · BSc Cybercrime & IT Security, First Class Honours
Recent Publications Recent Project ↗
About

From the Field to Threat Research

Originally from Thailand and now based in Ireland, my cybersecurity journey began in Incident Response at ReliaQuest. In the SOC, my days were defined by high-pressure environments, live triage, and containing active compromises. I spent countless hours digging through raw event logs, tracing lateral movement, and analyzing endpoint data to eject threat actors from enterprise networks. I learned an incredible amount from my peers and mentors in this role (Thank you SGA <3).

Witnessing these attacks firsthand taught me how attackers operate, but it also sparked a big shift in my perspective. Evicting an attacker solves today's emergency, but true defense requires understanding where they are going tomorrow. I wanted to move from a reactive posture of stopping the bleeding to a proactive one.

This realization drove my transition into Cyber Threat Intelligence.

What I Do Today

As a Threat Intelligence Researcher, I use my frontline IR experience to track the adversary for the wider security community. My work focuses on looking outside the enterprise perimeter to study the threat actors behind the keyboard, their financial motives, and their global infrastructure.

Operating under the broader threat research umbrella allows me to collaborate closely with our threat hunting and detection engineering teams. I am incredibly fortunate to work alongside brilliant people who inspire and push me forward every day. This environment allows me to contribute to a mission much larger than myself, protecting organizations on a global scale.

One of the most rewarding outcomes of this work is seeing original research I produce get shared and discussed across the industry, from major technical news vendors like BleepingComputer to global cybersecurity forums.

Life in CTI

⦾ Mapping out command-and-control (C2) networks, tracking bulletproof hosting providers, and identifying malicious domains before they are weaponized in active campaigns.
⦾ Researching underground forums, marketplaces, and ransomware-as-a-service (RaaS) leak sites to track Initial Access Brokers (IABs) and leaked credentials.
⦾ Studying the financial structures and monetization trends that fuel modern cybercriminal networks to help organizations anticipate attacks.

Outside the SOC

When I am offline and away from the screen, I teach myself how to work on cars after being inspired by my dad. When not under the hood I am collecting Pokemon cards and other various tidbits!

Recent Publications
001
Klue Integration Abused in Salesforce CRM Data TheftNEW
Attackers compromised a Klue competitive-intelligence service account, minted OAuth tokens, and walked straight into connected Salesforce environments. For nearly 24 hours they quietly pulled CRM records (account details, contacts, deal outcomes, pricing data) before dropping the patience act and firing off close to 1,000 REST API calls in a single 15-minute window, paging through results via Salesforce's QueryMore cursor from four distinct IP addresses. Jun 2026.
CRM SecurityOAuth AbuseAPI ExploitationData Exfiltration
002
Help-Desk Lures Drop KongTuke’s Evolved ModeloRAT
KongTuke abuses Microsoft Teams external federation to impersonate help-desk staff, using Unicode contact spoofing to look legitimate, then walks victims through downloading a portable Python environment via Dropbox and running a single pasted PowerShell command that kicks off the whole chain. Within five minutes, four persistence mechanisms lock in the ModeloRAT implant, including a SYSTEM-level scheduled task that standard remediation almost always misses, backed by three independent C2 paths built to survive a takedown. The group rotated through five Microsoft 365 tenants over 45 days to stay ahead of blocks. May 2026.
Malware AnalysisThreat IntelligenceInitial Access Brokers
003
DeepLoad Malware Pairs ClickFix Delivery with AI-Generated Evasion
DeepLoad arrives via ClickFix and buries its functional code beneath thousands of AI-generated meaningless variable assignments before running entirely in memory. It injects into Windows' LockAppHost.exe via APC injection while separately deploying a credential stealer and a malicious browser extension to capture stored passwords and live session tokens. A hidden WMI event subscription acts as a dead-man's switch, confirmed to re-execute the full attack chain three days after a cleaned host, and any connected USB drive gets seeded with 40+ disguised installer files. Mar 2026.
Malware AnalysisClickFixEvasion TechniquesWMI
004
What's Trending: Top Cyber Attacker Techniques, Sep–Nov 2025
Quarterly trend analysis for September–November 2025. BaoLoader led activity for the period, forging legitimacy through 26 real code-signing certificates built up across seven years. Infostealer detections dropped 69% after infrastructure takedowns, but Vidar 2.0 shifted focus to enterprise cloud credentials cached on personal devices to compensate. One CVE with a patch out since early 2025 still claimed 188 victims in October alone, and professional, scientific, and technical services firms saw a 57% spike in targeting. Jan 2026.
Threat TrendsMITRE ATT&CKRansomwareInitial Access
005
Storm-0249: Precision Endpoint Exploitation via EDR Abuse
Storm-0249 delivers a malicious MSI from a spoofed Microsoft domain that exploits Windows Installer's SYSTEM privileges to drop a trojanized DLL alongside a legitimate SentinelOne executable, hijacking the EDR agent via sideloading. All subsequent reconnaissance, C2 traffic, and MachineGuid extraction (used to bind ransomware encryption keys to the victim machine) then flow through that compromised security tool, looking identical to normal EDR activity. The approach works across EDR platforms and cuts the attacker's timeline from weeks to days. Dec 2025.
Threat Actor TrackingEDR AbuseDLL SideloadingClickFix
006
Speed, Scale, and Stealth: How Axios Powers Automated Phishing
Between June and August 2025, Axios user-agent activity surged 241%, with the HTTP library accounting for 24.44% of all flagged traffic as attackers leaned into its async workflows to run multi-stage credential theft in parallel. Campaigns embedded QR codes inside PDF attachments that sent victims to spoofed Outlook portals, where Axios middleware hijacked session tokens and MFA codes via the SAS BeginAuth mechanism. Paired with Microsoft Direct Send, the hit rate reached 70% versus 9.3% for traditional phishing, with lures built around PAYMENT, BONUS, and PAYSLIP subject lines. Sep 2025.
PhishingCredential TheftAutomationMicrosoft 365
007
First Look at CVE-2025-54309: Dissecting the Latest CrushFTP Exploit
CVE-2025-54309 is a CrushFTP AS2 validation flaw that opens a path to unauthenticated admin access through an alternative communication channel with weaker controls than the primary HTTPS endpoint. The key tell: server logs showed access denied while a crafted AS2 header had already granted full admin rights, which attackers used to enumerate the file system and attempt backdooring via setUserItem commands. A full RCE exploit appeared on dark web forums on July 22, 2025; IP allowlisting and policy controls stopped the final persistence step. Jul 2025.
CVE AnalysisVulnerability ResearchCrushFTPExploit Development