Skip to navigation

Threat reportMalwareTL-2026-2006

PATCHCORD, SHEETCORD & HACKERAI C2 Agent: New Malware Cluster Targets Afghan Telecom and South Asian Critical Infrastructure (APT36-linked)

highACTIVE

Acronis TRU identified three previously undocumented malware families — PATCHCORD (a C/C++ backdoor), SHEETCORD (a Go implant abusing the Google Sheets API for C2), and HACKERAI C2 Agent (a Go implant using GitHub Gists for C2, showing signs of AI-assisted development) — deployed in an active cyber-espionage campaign against Afghan Telecom (AFTEL), Afghanistan's Ministry of Communications and Information Technology, India's National Hydroelectric Power Corporation (NHPC), and Indian Ministry of Defence personnel.

Key facts

IdentifiersTL-2026-2006, CVE-2024-6387
SeverityHIGH
EPSSCVE-2024-6387: 99.5% (99.9th percentile)
Affected productsOpenSSH OpenSSH sshd; Microsoft Windows (desktop endpoints)
Weaknesses (CWE)CWE-362, CWE-364
AttributionAPT36 (medium confidence)
Nation-state nexusPakistan
MalwareHACKERAI C2 Agent, HackBrowserData, PATCHCORD, SHEETCORD, SHEETCREEP
ToolsAntnium, GateSentinel, Google Sheets API, HackBrowserData, Nim-based backdoor, SuperShell, metasploit
MITRE ATT&CK techniquesT1005 Data from Local System, T1027 Obfuscated Files or Information, T1036.005 Match Legitimate Resource Name or Location, T1041 Exfiltration Over C2 Channel, T1057 Process Discovery, T1059.001 Command and Scripting Interpreter, T1059.003 Windows Command Shell, T1071.001 Application Layer Protocol, T1082 System Information Discovery, T1102.002 Bidirectional Communication, T1105 Ingress Tool Transfer, T1140 Deobfuscate/Decode Files or Information and 17 more in the attack flow
MotivationESPIONAGE
Target sectorstelecoms, government administration, defense, energy, critical infrastructure
Target regionsafghanistan, South Asia, india
Detection rules9 (1 shown free)
Sources8 cited sources

Detection coverage

As of 2026-08-17, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-2006 across Splunk SPL, Microsoft KQL and Sigma, covering 48 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: Browser Shortcut Hijacking, VBScript Startup Persistence & Browser Credential Theft (PATCHCORD/SHEETCORD/HACKERAI Persistence)

Detects the browser-shortcut-hijacking persistence technique shared by PATCHCORD, SHEETCORD, and HACKERAI C2 Agent — rewriting Edge/Chrome/Firefox/Brave/Opera/Vivaldi .lnk targets across the Taskbar, Quick Launch, Start Menu, and Desktop to launch the implant before silently chain-loading the real browser — together with SHEETCORD's secondary VBScript Startup-folder persistence and HackBrowserData-driven browser credential harvesting.

  • Languages: Splunk SPL, Microsoft KQL, Sigma
  • ATT&CK: T1547.009, T1547.001, T1555.003
  • Data sources: sysmon, windows_security, edr_file_telemetry
  • Severity: high
  • Confidence: high

Splunk SPL

index=sysmon (EventCode=11 OR EventCode=1)
| eval activity=case(
    EventCode=11 AND match(TargetFilename,"(?i)\.lnk$") AND match(TargetFilename,"(?i)(edge|chrome|firefox|brave|opera|vivaldi)") AND match(TargetFilename,"(?i)(taskbar|quick launch|start menu|desktop)"), "browser_shortcut_hijack",
    EventCode=11 AND match(TargetFilename,"(?i)\\\\startup\\\\.*\.vbs$"), "vbscript_startup_persistence",
    EventCode=1 AND (match(Image,"(?i)hackbrowserdata") OR (match(CommandLine,"(?i)login data") AND match(CommandLine,"(?i)(chrome|edge|firefox)"))), "browser_credential_harvesting",
    true(), null())
| where isnotnull(activity)
| table _time host user activity Image TargetFilename CommandLine

Microsoft KQL

let ShortcutHijack = DeviceFileEvents
| where Timestamp > ago(24h)
| where FileName endswith ".lnk"
| where FolderPath has_any ("Taskbar","Quick Launch","Start Menu","Desktop")
| where FileName has_any ("Edge","Chrome","Firefox","Brave","Opera","Vivaldi")
| extend Activity = "browser_shortcut_hijack";
let VbsStartup = DeviceFileEvents
| where Timestamp > ago(24h)
| where FolderPath has "\\Startup\\" and FileName endswith ".vbs"
| extend Activity = "vbscript_startup_persistence";
let CredTheft = DeviceProcessEvents
| where Timestamp > ago(24h)
| where FileName has "HackBrowserData" or (ProcessCommandLine has "Login Data" and ProcessCommandLine has_any ("Chrome","Edge","Firefox"))
| extend Activity = "browser_credential_harvesting";
ShortcutHijack
| union VbsStartup, CredTheft
| project Timestamp, DeviceName, InitiatingProcessAccountName, Activity, FileName, FolderPath

Sigma

title: Browser Shortcut Hijack, VBScript Startup Persistence & Browser Credential Harvesting
id: 8f1c3a6e-4d2b-4e1a-9c7f-1a2b3c4d5e11
status: experimental
description: Detects rewriting of browser .lnk shortcut targets, VBScript loaders dropped into the Startup folder, and HackBrowserData-style credential harvesting, consistent with PATCHCORD/SHEETCORD/HACKERAI C2 Agent persistence and credential-access behavior.
references:
  - https://www.acronis.com/en/tru/posts/patchcord-new-malware-cluster-targets-afghan-telecom-and-south-asian-critical-infrastructure/
author: Threadlinqs Team
date: 2026/08/13
tags:
  - attack.persistence
  - attack.t1547.009
  - attack.t1547.001
  - attack.credential-access
  - attack.t1555.003
  - attack.tool.hackbrowserdata
  - attack.group.g0134
logsource:
  category: file_event
  product: windows
detection:
  selection_shortcut:
    TargetFilename|endswith: '.lnk'
    TargetFilename|contains:
      - 'Taskbar'
      - 'Quick Launch'
      - 'Start Menu'
      - 'Desktop'
  selection_shortcut_browser:
    TargetFilename|contains:
      - 'Edge'
      - 'Chrome'
      - 'Firefox'
      - 'Brave'
      - 'Opera'
      - 'Vivaldi'
  selection_vbs_startup:
    TargetFilename|contains: '\Startup\'
    TargetFilename|endswith: '.vbs'
  selection_credtheft:
    Image|contains: 'HackBrowserData'
  condition: (selection_shortcut and selection_shortcut_browser) or selection_vbs_startup or selection_credtheft
falsepositives:
  - Legitimate browser installer/updater re-pinning or repairing .lnk shortcuts during a version upgrade
  - IT-deployed VBScript logon/provisioning scripts placed in the Startup folder via GPO or SCCM
level: high

Known false positives:

  • Legitimate browser installer/updater re-pinning or repairing .lnk shortcuts during a version upgrade
  • IT-deployed VBScript logon/provisioning scripts placed in the Startup folder via GPO or SCCM

Tuning notes: Restrict the browser list to the six vendors named in the report (Edge, Chrome, Firefox, Brave, Opera, Vivaldi); baseline legitimate installer/updater processes permitted to rewrite these .lnk targets and exclude them by signed Authenticode publisher rather than by process name alone.

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)
  • Telecom Lure Installer Execution + Mass SSH Scanning (PATCHCORD/SHEETCORD Initial Access, CVE-2024-6387) · unlock with TL Creds
  • Browser Shortcut Hijacking, VBScript Startup Persistence & Browser Credential Theft (PATCHCORD/SHEETCORD/HACKERAI Persistence) · unlock with TL Creds
  • Living-off-Trusted-Services C2 via Google Sheets API / GitHub Gists + Config-Data Exfil (SHEETCORD/HACKERAI C2 Agent) · unlock with TL Creds
  • Telecom Lure Installer Execution + Mass SSH Scanning (PATCHCORD/SHEETCORD Initial Access, CVE-2024-6387) · unlock with TL Creds
  • Browser Shortcut Hijacking, VBScript Startup Persistence & Browser Credential Theft (PATCHCORD/SHEETCORD/HACKERAI Persistence) · unlock with TL Creds
  • Living-off-Trusted-Services C2 via Google Sheets API / GitHub Gists + Config-Data Exfil (SHEETCORD/HACKERAI C2 Agent) · unlock with TL Creds
  • Telecom Lure Installer Execution + Mass SSH Scanning (PATCHCORD/SHEETCORD Initial Access, CVE-2024-6387) · unlock with TL Creds
  • Living-off-Trusted-Services C2 via Google Sheets API / GitHub Gists + Config-Data Exfil (SHEETCORD/HACKERAI C2 Agent) · unlock with TL Creds

Indicators of compromise

As of 2026-08-17, Threadlinqs Intelligence tracks 48 indicators of compromise for TL-2026-2006.

Sample indicators (3 of 48)

  • IP address: 46.30.188[.]13
  • Domain: appstoore[.]solutions
  • File hash: 74d347785dc47f8cda3876826cdd3fb3935ac55dc8e9e0c0f96d5ef4e00089a2

Defanged, TLP:CLEAR. Full set of 48 indicators: unlock with TL Creds.

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

Technical analysis

Acronis TRU attributes the campaign with moderate confidence to APT36 (Transparent Tribe), a suspected Pakistan-linked group active since at least 2013.

Acronis's Threat Research Unit (researchers Subhajeet Singha, Darrel Virtusio, and Santiago Pontiroli) traced the cluster back to a suspicious ZIP archive uploaded to VirusTotal in June 2026 that referenced Afghan Telecom (AFTEL), Afghanistan's state-owned telecom operator. Pulling the thread exposed a live attacker staging server and three distinct, previously undocumented implants sharing infrastructure and delivery lures.

PATCHCORD is a compiled C/C++ backdoor distributed via Inno Setup installers that impersonate AFTEL VPN clients and telecom-management utilities. It performs host fingerprinting, executes attacker commands via cmd.exe, and establishes persistence by hijacking browser shortcuts. SHEETCORD is a Go-language evolution of that capability that routes command-and-control through the Google Sheets API using a hardcoded GCP service-account credential, giving the operators bidirectional C2 that rides on legitimate, TLS-encrypted Google infrastructure; Acronis TRU found significant code and functionality overlap between SHEETCORD and the previously reported SHEETCREEP implant (a C# Google-Sheets backdoor attributed with medium confidence to APT36 or a related Pakistan-linked subgroup by Zscaler ThreatLabz). A live SHEETCORD lure used a domain spoofing India's National Informatics Centre (NIC) to serve an installer disguised as a "Ministry of Defense Employee Breach Update," targeting Indian Ministry of Defence and Air Force personnel. HACKERAI C2 Agent is a third Go implant, first observed in the wild in September 2025, that instead abuses GitHub Gists as its C2 channel; its code contains debugging artifacts, redundant/duplicated XOR routines, hardcoded credentials, and clear-text debug strings consistent with rapid, AI-assisted ("vibeware") development rather than hand-written tradecraft — a production style Bitdefender separately documented as an emerging APT36 pattern in March 2026.

All three implants share a persistence technique Acronis calls browser-shortcut hijacking: the malware backs up the target's existing .lnk shortcuts for Edge, Chrome, Firefox, Brave, Opera, and Vivaldi across the taskbar, Quick Launch, Start Menu, and desktop, then rewrites each shortcut to launch the implant first before silently chain-loading the legitimate browser, keeping the victim unaware. SHEETCORD additionally drops a VBScript loader into the Windows Startup folder as a secondary persistence path. Execution is largely in-memory/reflective, minimizing on-disk forensic artifacts.

The operators' staging server, still active and reachable at the time of research, hosted a large offensive toolkit alongside the three implants: Metasploit, the Chinese-language SuperShell C2 panel, the Go/C GateSentinel C2 framework (co-located directly with PATCHCORD campaign files), the open-source Go C2 framework Antnium, HackBrowserData (an open-source browser credential harvester previously linked to APT36 operations), a rockyou.txt-derived password list, a Nim-based backdoor, and working exploit tooling for CVE-2024-6387 ("regreSSHion"), the OpenSSH signal-handler race condition, apparently staged for scanning and exploiting internet-facing OpenSSH instances. The server also held artifacts suggesting a broader collection scope beyond the three Windows implants documented here — an iOS call-history database schema and router configuration-backup archives — indicating the actors may also target mobile devices and network appliances in related operations, though Acronis TRU could not confirm those capabilities are wired into PATCHCORD/SHEETCORD/HACKERAI specifically. C2 domains sat behind Let's Encrypt TLS certificates that were rotated three times, each rotation covering a different bundled set of lookalike domains (impersonating AFTEL, Salaam Telecom, Afghanistan's Ministry of Communications and IT, and India's NHPC and Controller General of Defence Accounts), consistent with deliberate infrastructure cycling to evade domain/certificate-based blocklists. Acronis also identified a secondary server geolocated to Guangdong, China, hosting artifacts associated with the unrelated ValleyRAT/SilverFox malware family; TRU assessed the overlap as infrastructure co-location rather than evidence of operational coordination between the two actors.

Acronis TRU attributes the cluster to APT36 (Transparent Tribe) with moderate confidence, citing: sustained targeting of Afghan telecom and South Asian government/defense infrastructure consistent with APT36's historical scope; the presence of HackBrowserData, previously observed in APT36 tooling; co-location of the GateSentinel C2 framework with PATCHCORD campaign files; and SHEETCORD's close technical overlap with the Google-Sheets-based SHEETCREEP implant already linked to the group. The campaign reflects a broader shift, also documented independently by Bitdefender and Zscaler, of APT36 subgroups adopting "living-off-trusted-services" C2 (Google Sheets, GitHub Gists, Firebase, Slack, Discord) and AI-accelerated malware production to blend traffic into legitimate SaaS usage and increase implant throughput.

MITRE ATT&CK techniques by tactic:

Collection

T1005 Data from Local System; T1602.002 Network Device Configuration Dump

Defense Evasion

T1027 Obfuscated Files or Information; T1036.005 Match Legitimate Resource Name or Location; T1140 Deobfuscate/Decode Files or Information; T1620 Reflective Code Loading; T1622 Debugger Evasion

Exfiltration

T1041 Exfiltration Over C2 Channel

Discovery

T1057 Process Discovery; T1082 System Information Discovery

Execution

T1059.001 Command and Scripting Interpreter; T1059.003 Windows Command Shell; T1204.002 Malicious File

Command and Control

T1071.001 Application Layer Protocol; T1102.002 Bidirectional Communication; T1105 Ingress Tool Transfer

Initial Access

T1190 Exploit Public-Facing Application; T1566.001 Phishing; T1566.002 Spearphishing Link

Persistence

T1547.001 Registry Run Keys / Startup Folder; T1547.009 Shortcut Modification

Credential Access

T1555.003 Credentials from Web Browsers

Resource Development

T1583.001 Domains; T1583.006 Acquire Infrastructure; T1587.001 Develop Capabilities; T1587.003 Digital Certificates; T1588.002 Tool; T1588.005 Obtain Capability: Exploits

Reconnaissance

T1595.002 Vulnerability Scanning

Remediation

Patches

  • Upgrade OpenSSH to version 9.8p1 or later to remediate CVE-2024-6387.

Immediate actions

  • Block or alert on outbound traffic to the Google Sheets API and GitHub Gists/GitHub API from endpoints in telecom, government, and defense environments where those services are not an approved business dependency.
  • Audit browser shortcut (.lnk) target paths on Windows endpoints for Edge, Chrome, Firefox, Brave, Opera, and Vivaldi for unauthorized redirection to unexpected executables.
  • Verify authenticity of any VPN client or telecom-management installer purporting to originate from Afghan Telecom (AFTEL) or Salaam Telecom before execution; block known impersonating domains and installer hashes once published in the Acronis IOC set.
  • Patch or upgrade OpenSSH to 9.8p1 or later on all internet-facing hosts to close CVE-2024-6387 (regreSSHion).

Workarounds

  • Interim mitigation for CVE-2024-6387: set LoginGraceTime to 0 in sshd_config (note: this increases exposure to a related denial-of-service condition and is not a substitute for patching).

Longer-term hardening

  • Deploy EDR/behavioral detection tuned for in-memory and reflective code loading, since these implants minimize on-disk artifacts.
  • Apply CASB/TLS-inspection controls to legitimate SaaS platforms (Google Sheets, GitHub, Firebase, Slack, Discord) that are increasingly abused as C2 transport by APT36-linked subgroups.
  • Restrict write access to the Windows Startup folder and alert on new VBScript (.vbs) file creation there.
  • Monitor for and rotate any exposed GCP service-account credentials; enforce least-privilege scoping on cloud service accounts used by internal automation to reduce reuse risk if credentials are harvested.

Affected and fixed versions

  • OpenSSH — OpenSSH sshd
    Vulnerable versions: 8.5p1-9.7p1
    Fixed in: 9.8p1 and later
  • Microsoft — Windows (desktop endpoints)
    Vulnerable versions: PATCHCORD, SHEETCORD, and HACKERAI C2 Agent are Windows-targeted implants; specific build numbers not specified in reporting

Timeline

Last updated · 2 updates · revalidated 2× · latest source

  • APT36 (Transparent Tribe) first tracked as an active, Pakistan-linked threat group targeting South Asian government, military, and diplomatic organizations.
  • CVE-2024-6387 ("regreSSHion"), an OpenSSH sshd signal-handler race condition enabling unauthenticated RCE, is publicly disclosed; exploit tooling for it is later found staged on the actor's server.
  • Earliest observed domain/infrastructure associated with the HACKERAI C2 Agent (GitHub Gists-based implant) becomes active.
  • PATCHCORD/SHEETCORD/HACKERAI campaign infrastructure confirmed active, including domains impersonating AFTEL and related South Asian government/telecom entities.
  • Actor registers the locked-ioc-12 domain, later hardcoded as PATCHCORD's primary C2 domain.
  • Acronis TRU telemetry places the PATCHCORD backdoor as operational since at least March 2026, initially focused on Afghan telecom providers.
  • Bitdefender publishes "APT36: A Nightmare of Vibeware," documenting the group's shift to AI-assisted, high-volume ("vibeware") malware production and abuse of Google Sheets, Slack, Discord, and Firebase for C2 — a pattern later mirrored in HACKERAI C2 Agent.
  • Lure domain locked-ioc-16 first observed registered/active.
  • Lure domain locked-ioc-14, impersonating Afghan Telecom (AFTEL), first observed active.
  • Lure domain locked-ioc-22, impersonating India's National Informatics Centre, first observed active.
  • Acronis TRU discovers a suspicious ZIP archive on VirusTotal referencing Afghan Telecom (AFTEL), triggering the investigation that uncovers PATCHCORD, SHEETCORD, and HACKERAI C2 Agent.
  • Acronis TRU (Subhajeet Singha, Darrel Virtusio, Santiago Pontiroli) publishes findings, attributing the campaign with moderate confidence to APT36 (Transparent Tribe).
  • The locked-ioc-12/locked-ioc-24 C2 infrastructure confirmed still live and reachable.
  • Wider public pickup of the Acronis TRU research by Security Affairs, The Hacker News, GBHackers, and CyberSecurityNews, with additional technical detail (C2 domain/IP, lure domains, sample hash, registry persistence value).

Update history

Sources cited

More from APT36

Previous and next in malware

Drafted by the Threadlinqs research pipeline from 8 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