A complex wave of credential theft and espionage activity has been identified across five distinct OTX pulses. The primary threats involve the FortiBleed campaign, which utilizes a "CyberStrike Harvester" to compromise FortiGate SSL VPN credentials globally through password spraying and offline cracking. Simultaneously, the TAG-195 threat actor (Golden Chickens) has expanded its Malware-as-a-Service (MaaS) ecosystem with modular tools like TinyEgg and ChonkyChicken, focusing on browser credential theft and WebSocket C2 communications.
In the espionage sector, TA458 (Russia-aligned) is exploiting half-click zero-day vulnerabilities in webmail platforms (SOGo, Zimbra) to deliver SpyPress malware, specifically targeting Eastern European government entities. Additionally, a supply chain attack via the ClawHub AI marketplace has introduced macOS infostealers (AMOS, ClawHavoc) into enterprise environments, while the Hades implant and Hermes AI Agent were used in a targeted attack against Thailand's Ministry of Finance. The collective objective is the mass exfiltration of credentials, session cookies, and sensitive government data.
Threat Actor / Malware Profile
CyberStrike Harvester (FortiBleed)
- Actor: Unknown (Large-scale credential theft operation)
- Distribution: Credential stuffing and password spraying against internet-facing FortiGate SSL VPNs.
- Behavior: Harvests configuration files, performs offline cracking using hashcat, and utilizes the EKZ Infostealer component.
- C2: Exfiltrates captured credentials via file transfer mechanisms post-authentication.
TAG-195 (Golden Chickens / Venom Spider)
- Actor: Financially motivated MaaS developer.
- Families: TinyEgg (Backdoor), ChonkyChicken (Browser Stealer), TerraStealerV2.
- Behavior: TinyEgg establishes initial access, loading ChonkyChicken to steal browser credentials and automate sessions. Uses WebSocket protocols for C2 to evade detection.
TA458 (Operation RoundPress)
- Actor: Russia-aligned, likely GRU.
- Malware: SpyPress.
- Behavior: Exploits "half-click" XSS zero-days in webmail (CVE-2026-8496, CVE-2025-27915). Requires no user interaction beyond opening an email.
Hermes AI Agent & Hades
- Actor: Unknown (Espionage)
- Behavior: Hermes operates autonomously ("YOLO mode") to stage the custom Go-based Hades implant. Exploits vulnerabilities in Apache Hadoop and web servers.
IOC Analysis
The provided indicators reveal a multi-vector approach:
- File Hashes: A significant volume of SHA256 hashes (e.g.,
2758f4d71a2a2dfdefab81737c2d776b2a3dafe5844fdd2157e089a28447ca98,6c23b7723a9f69ea48f02c8fe13fd60ecbfc2fb28e32e481c46ec968a66c66cd) correspond to the CyberStrike Harvester, TAG-195 payloads, and AI supply chain malware. SOC teams should prioritize scanning endpoints and cloud storage for these hashes. - CVEs: Critical vulnerabilities including CVE-2026-35616 (FortiBleed), CVE-2026-8496 (SOGo), and CVE-2026-43500 (HiveServer2) are key exploit vectors. Vulnerability scanning must be updated to include these 2026 CVEs.
- Domains & IPs: Infrastructure includes
43.246.208.207(Thailand MoF C2), and domainsxtrafftrck.net,screenly.cam,ahdaratlegalservices.com(TAG-195), andupgybj.store(TA458). These should be blocked at the perimeter and DNS layers. - Tooling: Decrypting and analyzing these IOCs requires tools like VirusTotal for hash reputation, Cortex for analyzer execution, and MISP for correlation.
Detection Engineering
title: Potential Webshell Activity via Hades/VShell
id: 6f3a9b8d-5c1e-4a2d-8b9e-1f3c5d6a7b8c
description: Detects potential webshell activity associated with Hades implant or VShell as seen in Thailand MoF attacks, characterized by suspicious process spawning from web server services.
status: stable
author: Security Arsenal
date: 2026/07/24
references:
- https://hunt.io/blog/thailand-ministry-finance-targeted-with-hermes-ai-agent
tags:
- attack.persistence
- attack.t1505.003
logsource:
category: process_creation
product: windows
detection:
selection:
ParentImage|endswith:
- '\w3wp.exe'
- '\httpd.exe'
- '\nginx.exe'
Image|endswith:
- '\cmd.exe'
- '\powershell.exe'
- '\pwsh.exe'
condition: selection
falsepositives:
- Legitimate administrative scripts
level: high
---
title: Suspicious PowerShell Loader - TAG-195 TinyEgg/ChonkyChicken
id: 1a2b3c4d-5e6f-7a8b-9c0d-1e2f3a4b5c6d
description: Detects PowerShell download cradles and obfuscation often used by MaaS families like TinyEgg and ChonkyChicken to load modular stealers.
status: stable
author: Security Arsenal
date: 2026/07/24
references:
- https://www.recordedfuture.com/research/tag-195-evolves-maas-ecosystem
tags:
- attack.execution
- attack.t1059.001
logsource:
category: process_creation
product: windows
detection:
selection:
Image|endswith: '\powershell.exe'
CommandLine|contains:
- 'DownloadString'
- 'IEX'
- 'Invoke-Expression'
- 'FromBase64String'
filter:
CommandLine|contains:
- 'windowsupdate'
- 'microsoft'
condition: selection and not filter
falsepositives:
- System administration scripts
level: medium
---
title: FortiBleed Credential Cracking Tool Execution
id: 9e8f7a6b-5c4d-3e2f-1a0b-9c8d7e6f5a4b
description: Detects execution of credential cracking tools like hashcat often used post-harvest by CyberStrike Harvester to process stolen FortiGate credentials.
status: stable
author: Security Arsenal
date: 2026/07/24
references:
- https://arcticwolf.com/resources/blog/inside-fortibleed-reverse-engineering-the-cyberstrike-harvester-behind-a-global-fortigate-credential-factory/
tags:
- attack.credential_access
- attack.t1110.003
logsource:
category: process_creation
product: windows
detection:
selection:
Image|endswith: '\hashcat.exe'
or OriginalFileName: 'hashcat.exe'
condition: selection
falsepositives:
- Authorized penetration testing
level: critical
kql
// Hunt for TAG-195 C2 Domains and TA458 Infrastructure
// Searches for network connections to known malicious domains and IPs
DeviceNetworkEvents
| where Timestamp > ago(7d)
| where RemoteUrl in~ (
"xtrafftrck.net",
"screenly.cam",
"ahdaratlegalservices.com",
"upgybj.store",
"xsza.net",
"laosji.net"
)
or RemoteIP == "43.246.208.207"
| summarize count() by DeviceName, RemoteUrl, RemoteIP, InitiatingProcessAccountName
| order by count_ desc
powershell
# IOC Hunt Script for July 24, 2026 OTX Pulses
# Checks for malicious file hashes and C2 domain presence
$MaliciousHashes = @(
"7f74bb6ba185978134c318bc5f91d23c", # CyberStrike/EKZ
"2758f4d71a2a2dfdefab81737c2d776b2a3dafe5844fdd2157e089a28447ca98", # CyberStrike
"6c23b7723a9f69ea48f02c8fe13fd60ecbfc2fb28e32e481c46ec968a66c66cd", # TinyEgg
"45c8cbaeb5c7708e7b8030e701747c65203958e82eddc41f39e0ca93bd36c114", # ChonkyChicken
"b7b322f4638ead5c39031ffc7ca8c791c8d47211b09449f7ceb49f0c32a19b45", # ChonkyChicken
"818aea6143282b352fdfdc0f3ebf77a36e54eb3befb5cad1a355a99ab97c6aa7" # AMOS/Cluw
)
$C2Domains = @(
"xtrafftrck.net",
"screenly.cam",
"ahdaratlegalservices.com",
"laosji.net",
"upgybj.store",
"xsza.net"
)
Write-Host "[+] Scanning running processes for known malicious hashes..." -ForegroundColor Cyan
Get-Process | ForEach-Object {
$proc = $_
try {
$path = $proc.Path
if ($path -and (Test-Path $path)) {
$hash = (Get-FileHash -Path $path -Algorithm MD5).Hash.ToLower()
if ($MaliciousHashes -contains $hash) {
Write-Host "[!] MALICIOUS PROCESS DETECTED: PID $($proc.Id) - $($proc.ProcessName) - $path" -ForegroundColor Red
}
}
} catch {}
}
Write-Host "[+] Checking DNS Client Cache for C2 domains..." -ForegroundColor Cyan
Get-DnsClientCache | Where-Object { $C2Domains -contains $_.Entry } | ForEach-Object {
Write-Host "[!] SUSPICIOUS DNS ENTRY FOUND: $($_.Entry) -> $($_.Data)" -ForegroundColor Yellow
}
# Response Priorities
* **Immediate:**
* Block all listed domains (`xtrafftrck.net`, `upgybj.store`, etc.) and IP addresses (`43.246.208.207`) at the firewall and proxy.
* Scan all endpoints for the provided file hashes (CyberStrike, TinyEgg, AMOS).
* Initiate hunting for webshell activity on public-facing servers (specifically Apache and IIS).
* **24 Hours:**
* Verify user identities for accounts with access to FortiGate SSL VPNs and webmail platforms; reset credentials if suspicious activity is detected.
* Patch vulnerabilities associated with CVE-2026-35616 (FortiBleed) and CVE-2026-8496 (SOGo) if patches are available, or apply mitigations (WAF rules).
* Review AI agent marketplace access (ClawHub) and quarantine any downloaded skills matching the SHA256s provided.
* **1 Week:**
* Architecture hardening: Disable password spraying attacks by implementing MFA on all VPN and webmail entry points.
* Review and restrict the use of AI coding assistants/skills within the development environment.
* Conduct a forensic review of the Thailand MoF campaign indicators (Hades/Hermes) to ensure no lateral movement occurred if the organization operates in the targeted sectors.
Related Resources
Security Arsenal Incident Response Managed SOC & MDR Services AlertMonitor Threat Detection From The Dark Side Intel Hub
Is your security operations ready?
Get a free SOC assessment or see how AlertMonitor cuts through alert noise with automated triage.