Skip to navigation

Threat reportVulnerabilityTL-2026-0264

Craft CMS Zero-Day RCE Exploitation by Mimo Group (CVE-2025-32432, CVSS 10.0)

criticalMONITORING

Critical unauthenticated remote code execution vulnerability in Craft CMS (CVE-2025-32432, CVSS 10.0) actively exploited since February 2025 by the Mimo threat group to deploy cryptocurrency miners, residential proxyware, and web shells. Chained with Yii framework flaw CVE-2024-58136 for full server compromise.

Key facts

IdentifiersTL-2026-0264, CVE-2025-32432, CVE-2024-58136
Also known asOperation Mimolette, Mimo Craft CMS Campaign
SeverityCRITICAL
CVSS10 (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:L)
Exploitation statusLinked CVE exploited in the wild; Listed in CISA KEV: CVE-2025-32432 (added ), CVE-2024-58136 (added )
EPSSCVE-2025-32432: 99.7% (99.9th percentile); CVE-2024-58136: 87.7% (99.7th percentile)
Affected productsCraft CMS Craft CMS; Yii Software Yii Framework 2
Weaknesses (CWE)CWE-94
AttributionHezb (high confidence)
Nation-state nexusTurkey
MalwareIPRoyal Pawns, Mimo Loader, xmrig
ToolsGSocket, UPX
MITRE ATT&CK techniquesT1018 Remote System Discovery, T1021 Remote Services, T1027 Obfuscated Files or Information, T1053 Scheduled Task/Job, T1057 Process Discovery, T1059 Command and Scripting Interpreter, T1071 Application Layer Protocol, T1090 Proxy, T1190 Exploit Public-Facing Application, T1203 Exploitation for Client Execution, T1496 Resource Hijacking, T1505 Server Software Component and 6 more in the attack flow
MotivationFINANCIAL
Target sectorstechnology, media, e-commerce, education, government, small-business, web-hosting
Target regionsNorth America, Europe, Global
Detection rules9 (1 shown free)
Sources12 cited sources

Detection coverage

As of 2026-03-21, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-0264 across Splunk SPL, Microsoft KQL and Sigma, covering 30 indicator(s) of compromise. The whole corpus is readable without an account. Every threat comes with one complete detection in Splunk SPL, Microsoft KQL and Sigma, free for everyone; a free account adds daily TL Creds to unlock the rest. Paid plans add a larger daily allowance, correlation and the MCP server. Threadlinqs MCP server · View plans.

Free detection: Mimo 4l4md4r Loader Execution and LD_PRELOAD Rootkit Installation

Detects Mimo/Hezb post-exploitation activity: the 4l4md4r.sh dropper script download and execution, the 4l4md4r Go-based loader binary execution, and the installation of the alamdar.so LD_PRELOAD rootkit via modification of /etc/ld.so.preload. This covers the core Mimo persistence and defense evasion mechanism where the rootkit hooks readdir/readdir64 to hide malicious processes.

  • Languages: Splunk SPL, Microsoft KQL, Sigma
  • Severity: critical
  • Confidence: high

Splunk SPL

index=sysmon sourcetype="XmlWinEventLog:Microsoft-Windows-Sysmon/Operational"
(EventCode=1 OR EventCode=11)
(
    (EventCode=1 AND (
        (CommandLine="*4l4md4r*")
        OR (CommandLine="*alamdar*")
        OR (CommandLine="*curl*" AND CommandLine="*4l4md4r.sh*")
        OR (CommandLine="*wget*" AND CommandLine="*4l4md4r.sh*")
        OR (CommandLine="*python*" AND CommandLine="*urllib*" AND CommandLine="*fbi*")
        OR (Image="*4l4md4r*")
    ))
    OR (EventCode=11 AND (
        TargetFilename="/etc/ld.so.preload"
        OR TargetFilename="*alamdar.so"
        OR TargetFilename="*4l4md4r*"
        OR TargetFilename="*hezb.x86_64*"
    ))
)
| eval artifact_type=case(
    match(CommandLine, "4l4md4r\.sh"), "dropper_download",
    match(Image, "4l4md4r"), "loader_execution",
    match(TargetFilename, "ld\.so\.preload"), "rootkit_install",
    match(TargetFilename, "alamdar\.so"), "rootkit_library",
    match(TargetFilename, "hezb"), "proxyware_drop",
    match(CommandLine, "alamdar"), "miner_execution",
    1=1, "unknown"
)
| stats count AS event_count, values(artifact_type) AS artifacts, values(CommandLine) AS commands, values(TargetFilename) AS files BY Computer, User
| table _time, Computer, User, artifacts, commands, files, event_count

Microsoft KQL

union DeviceProcessEvents, DeviceFileEvents
| where Timestamp > ago(24h)
| where
    (ActionType == "ProcessCreated" and (
        ProcessCommandLine has "4l4md4r"
        or ProcessCommandLine has "alamdar"
        or (ProcessCommandLine has "curl" and ProcessCommandLine has "4l4md4r.sh")
        or (ProcessCommandLine has "wget" and ProcessCommandLine has "4l4md4r.sh")
        or (ProcessCommandLine has "python" and ProcessCommandLine has "urllib" and ProcessCommandLine has "fbi")
        or FileName == "4l4md4r"
    ))
    or (ActionType in ("FileCreated", "FileModified") and (
        FolderPath == "/etc/ld.so.preload"
        or FolderPath endswith "alamdar.so"
        or FolderPath has "4l4md4r"
        or FolderPath has "hezb.x86_64"
    ))
| extend ArtifactType = case(
    ProcessCommandLine has "4l4md4r.sh" and ProcessCommandLine has_any ("curl", "wget"), "dropper_download",
    FileName == "4l4md4r", "loader_execution",
    FolderPath == "/etc/ld.so.preload", "rootkit_install",
    FolderPath endswith "alamdar.so", "rootkit_library",
    FolderPath has "hezb", "proxyware_drop",
    ProcessCommandLine has "alamdar", "miner_execution",
    true, "unknown"
)
| summarize EventCount = count(), Artifacts = make_set(ArtifactType), Commands = make_set(ProcessCommandLine), Files = make_set(FolderPath) by DeviceName, AccountName
| project Timestamp = now(), DeviceName, AccountName, Artifacts, Commands, Files, EventCount

Sigma

title: Mimo 4l4md4r Loader and LD_PRELOAD Rootkit Installation
id: d4e5f6a7-b8c9-4d0e-1f2a-3b4c5d6e7f80
status: experimental
description: Detects the Mimo/Hezb threat group post-exploitation chain including the 4l4md4r.sh dropper download, 4l4md4r loader execution, and alamdar.so rootkit installation via /etc/ld.so.preload modification.
references:
    - https://blog.sekoia.io/the-sharp-taste-of-mimolette-analyzing-mimos-latest-campaign-targeting-craft-cms/
    - https://securitylabs.datadoghq.com/articles/beyond-mimolette-tracking-mimo-expansion-magento-cms-docker/
author: Threadlinqs Team
date: 2026-03-21
tags:
    - attack.persistence
    - attack.t1574.006
    - attack.defense_evasion
    - attack.t1547.004
    - attack.execution
    - attack.t1059.004
    - cve.2025.32432
logsource:
    category: process_creation
    product: linux
detection:
    selection_dropper:
        CommandLine|contains:
            - '4l4md4r.sh'
            - '4l4md4r'
    selection_download:
        CommandLine|contains:
            - 'curl'
            - 'wget'
        CommandLine|contains:
            - '4l4md4r'
    selection_loader:
        Image|endswith:
            - '/4l4md4r'
    selection_python_alias:
        CommandLine|contains|all:
            - 'python'
            - 'urllib'
            - 'fbi'
    condition: selection_dropper or selection_download or selection_loader or selection_python_alias
falsepositives:
    - Very unlikely — 4l4md4r and alamdar are unique Mimo malware identifiers with no legitimate use
level: critical

Tuning notes: This is a high-fidelity detection based on unique Mimo IOCs. No tuning should be needed. Consider adding SHA256 hash matching for the known Mimo binaries (bf2727a50a1c083fe478de1e51bded8a8eea6d6f1d02855161df8be42965f976 for 4l4md4r, ef108d6f7d9f2fe240a32f2cde594d3e54319237cedd21d13ab5e10ad4d05dbb for alamdar.so) if your SIEM ingests file hash events.

8 more detection rules (Splunk SPL, Microsoft KQL, Sigma) · unlock with TL Creds. Sign up free to unlock the detection rules or query them with the Threadlinqs MCP server (Purple)
  • Craft CMS CVE-2025-32432 Exploitation — Web Server Child Process Spawning · unlock with TL Creds
  • Mimo Web Shell Deployment and Cron-Based Persistence · unlock with TL Creds
  • Mimo C2 Infrastructure Connection and SSH Lateral Movement · unlock with TL Creds
  • Craft CMS PHP Session Poisoning and Object Injection — CVE-2025-32432 · unlock with TL Creds
  • LD_PRELOAD Rootkit Installation and Process Hiding — Mimo [indicator] · unlock with TL Creds
  • Mimo XMRig Cryptominer and IPRoyal Proxyware Execution · unlock with TL Creds
  • Craft CMS CVE-2025-32432 RCE Exploitation — HTTP Request Detection · unlock with TL Creds
  • Mimo GSocket C2 Communication and Cryptomining Activity · unlock with TL Creds

Indicators of compromise

As of 2026-03-21, Threadlinqs Intelligence tracks 30 indicators of compromise for TL-2026-0264.

Sample indicators (3 of 30)

  • IP address: 85.106.113[.]168
  • File hash: bf2727a50a1c083fe478de1e51bded8a8eea6d6f1d02855161df8be42965f976
  • IP address: 109.205.213[.]203

Defanged, TLP:CLEAR. Full set of 30 indicators: unlock with TL Creds.

30 indicators of compromise · unlock with TL Creds. Compare plans

Technical analysis

Added to CISA KEV March 20, 2026 with remediation deadline April 3, 2026.

CVE-2025-32432 is a maximum-severity code injection vulnerability (CWE-94) in Craft CMS affecting versions 3.0.0-RC1 through 3.9.14, 4.0.0-RC1 through 4.14.14, and 5.0.0-RC1 through 5.6.16. The flaw resides in the asset transform generation feature, specifically the /index.php?p=actions/assets/generate-transform endpoint, and enables unauthenticated remote code execution.

The exploitation chain operates in three stages. First, an attacker plants malicious PHP code by sending a crafted GET request with a return URL parameter containing PHP code (e.g., <?=exec($_GET[''cmd'']);die()?>), which Craft CMS saves into a server-side PHP session file at /var/lib/php/sessions/sess_[SESSION_ID]. Second, the attacker exploits the __class bypass in the image transform endpoint by sending a POST request with a malicious JSON payload containing a specially crafted field prefix (''as hack'') that triggers CVE-2024-58136, an input validation flaw in the Yii 2 framework (versions before 2.0.52). This loads the PhpManager gadget and points it to the poisoned session file. Third, when PhpManager loads the session file, the attacker''s PHP payload executes, granting full remote code execution.

The vulnerability was discovered by Orange Cyberdefense SensePost and first observed being exploited in the wild in mid-February 2025, approximately two months before public disclosure on April 10, 2025. By mid-April 2025, approximately 13,000 vulnerable Craft CMS instances were identified globally, with nearly 300 confirmed compromised.

The Mimo threat group (also known as Hezb), a financially motivated intrusion set believed to be physically located in Turkey and active since March 2022, has been the primary exploiter of this vulnerability. Mimo''s attack chain deploys a PHP web shell for persistent access, then downloads and executes a shell script named ''4l4md4r.sh'' using curl, wget, or Python''s urllib2 library (distinctively aliased as ''fbi''). The script deploys the 4l4md4r Go-based loader, which installs the alamdar.so LD_PRELOAD rootkit for process hiding, the IPRoyal residential proxyware (hezb.x86_64) for bandwidth monetization, and the XMRig cryptocurrency miner for Monero mining via MoneroOcean and C3Pool.

Mimo has a history of exploiting high-profile vulnerabilities including CVE-2021-44228 (Log4j), CVE-2022-26134 (Confluence), CVE-2023-27350 (PaperCut), and CVE-2023-46604 (ActiveMQ). The group has also deployed the Mimus ransomware, a Go-based fork of MauriCrypt. The campaign has expanded beyond Craft CMS to target Magento and Docker environments.

CISA added CVE-2025-32432 to its Known Exploited Vulnerabilities catalog on March 20, 2026, requiring federal agencies to apply mitigations by April 3, 2026.

MITRE ATT&CK techniques by tactic:

discovery

T1018 Remote System Discovery; T1057 Process Discovery

lateral-movement

T1021 Remote Services

defense-evasion

T1027 Obfuscated Files or Information; T1564 Hide Artifacts

execution

T1053 Scheduled Task/Job; T1059 Command and Scripting Interpreter; T1203 Exploitation for Client Execution

command-and-control

T1071 Application Layer Protocol; T1090 Proxy

initial-access

T1190 Exploit Public-Facing Application

impact

T1496 Resource Hijacking

persistence

T1505 Server Software Component; T1547 Boot or Logon Autostart Execution; T1574 Hijack Execution Flow

privilege-escalation

T1548 Abuse Elevation Control Mechanism

credential-access

T1552 Unsecured Credentials

resource-development

T1583 Acquire Infrastructure

Remediation

Patches

  • Craft CMS 3.9.15 (patches CVE-2025-32432 for 3.x branch)
  • Craft CMS 4.14.15 (patches CVE-2025-32432 for 4.x branch)
  • Craft CMS 5.6.17 (patches CVE-2025-32432 for 5.x branch)
  • Yii Framework 2.0.52 (patches CVE-2024-58136)

Immediate actions

  • Update Craft CMS to patched versions: 3.9.15, 4.14.15, or 5.6.17 immediately
  • Update Yii framework to version 2.0.52 or later
  • Block POST requests to /index.php?p=actions/assets/generate-transform containing __class in the body at WAF level
  • Search web server logs for suspicious POST requests to actions/assets/generate-transform endpoint
  • Check /etc/ld.so.preload for unauthorized entries referencing alamdar.so
  • Scan for processes named 4l4md4r, alamdar, or hezb.x86_64
  • Block known C2 IPs: 85.106.113.168, 109.205.213.203, 193.32.162.10, 15.188.246.198

Workarounds

  • If patching is not immediately possible, disable the asset transform generation endpoint
  • Restrict access to Craft CMS admin endpoints via network-level controls
  • Implement strict Content-Security-Policy headers to limit code execution
  • Discontinue use of Craft CMS if mitigations are unavailable per CISA BOD 22-01 guidance

Longer-term hardening

  • Implement Web Application Firewall rules to detect PHP object injection patterns
  • Deploy endpoint detection for LD_PRELOAD hijacking and rootkit installation
  • Monitor for unauthorized cryptocurrency mining activity via CPU usage anomalies
  • Implement file integrity monitoring on /etc/ld.so.preload and PHP session directories
  • Establish regular vulnerability scanning for CMS platforms
  • Monitor for GSocket-based C2 communications

Affected and fixed versions

  • Craft CMS — Craft CMS
    Vulnerable versions: 3.0.0-RC1 through 3.9.14; 4.0.0-RC1 through 4.14.14; 5.0.0-RC1 through 5.6.16
    Fixed in: 3.9.15; 4.14.15; 5.6.17
  • Yii Software — Yii Framework 2
    Vulnerable versions: < 2.0.52
    Fixed in: 2.0.52

Timeline

  • Mimo threat group first observed active, initially exploiting Log4j (CVE-2021-44228) and Confluence (CVE-2022-26134) vulnerabilities
  • CVE-2025-32432 first exploited in the wild as a zero-day by unknown threat actors, approximately two months before public disclosure
  • Exploitation attempts detected on Sekoia honeypot infrastructure, beginning sustained observation period through May 2025
  • Craft CMS releases patches in versions 3.9.15, 4.14.15, and 5.6.17 addressing CVE-2025-32432. Yii Framework 2.0.52 patches CVE-2024-58136
  • Full vulnerability details publicly disclosed. Censys identifies approximately 78,984 Craft CMS instances, with nearly 13,000 running vulnerable versions
  • GitHub Advisory GHSA-f3gw-9ww9-jmc3 published with full technical details. Approximately 300 servers confirmed compromised
  • End of sustained exploitation observation period on Sekoia honeypot, covering multiple Mimo intrusion attempts
  • Mimo C2 infrastructure active at locked-ioc-11:21, secondary C2 at locked-ioc-16:21 active May 28-30
  • Sekoia publishes detailed analysis The Sharp Taste of Mimolette documenting Mimo complete Craft CMS exploitation campaign
  • Mimo observed using locked-ioc-15:80 for payload hosting, expanding infrastructure
  • Datadog Security Labs documents Mimo expansion beyond Craft CMS to Magento and Docker environments
  • CISA adds CVE-2025-32432 to Known Exploited Vulnerabilities catalog, orders federal agencies to patch by April 3, 2026
  • CISA BOD 22-01 federal remediation deadline for CVE-2025-32432
  • As of 2026-05-29, CVE-2025-32432 (Craft CMS, CVSS 10.0) is patched but remains actively exploited per its March 20, 2026 CISA KEV listing (deadline Apr 3, 2026), and the Mimo/Hezb actor shows no disruption. With ongoing exploitation of unpatched instances and the actor expanding to Magento/Docker, the threat stays MONITORING.

Sources cited

Other reports on CVE-2025-32432

Previous and next in vulnerability

Drafted by the Threadlinqs research pipeline from 12 sources; published ; last updated —.

Threadlinqs Intelligence — Real-Time Threat Detection Platform

// threat_feed
$ sort --newest
Showing all threats

Live intelligence console

Threat level

Latest Threats