Skip to navigation

Threat reportThreat IntelligenceTL-2026-4359

Linux clear-text credential harvesting techniques (process memory, PAM backdoors, bash history, ptrace tooling)

mediumMONITORING

DFIR.ch (Stephan Berger) documents how threat actors collect clear-text passwords on compromised Linux systems via shell history, /proc and process memory, ptrace-based interception (3snake), PAM manipulation, and backdoored edge-device login portals, then use them for lateral movement and cloud pivoting. First disclosed .

Key facts

IdentifiersTL-2026-4359
SeverityMEDIUM
First disclosed
Affected productsLinux Linux servers (PAM, sshd, sudo, /proc); Citrix NetScaler ADC / Gateway authentication portal (backdoored login page example)
Weaknesses (CWE)CWE-522, CWE-312
MalwareDecoy Dog RAT
Tools3snake, pspy, truffleproc
MITRE ATT&CK techniquesT1003.007 Proc Filesystem, T1021.004 SSH, T1055.008 Ptrace System Calls, T1056.001 Keylogging, T1056.003 Web Portal Capture, T1057 Process Discovery, T1078 Valid Accounts, T1078.004 Cloud Accounts, T1190 Exploit Public-Facing Application, T1552.001 Credentials In Files, T1552.003 Shell History, T1556.003 Pluggable Authentication Modules
MotivationUNKNOWN
Target sectorstechnology, telecoms, government administration
Target regionsGlobal
Detection rules9 (1 shown free)
Sources11 cited sources

Detection coverage

As of 2026-10-24, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-4359 across Splunk SPL, Microsoft KQL and Sigma, covering 16 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: Linux Ptrace Attach Credential Interception (3snake / truffleproc / gdb) - Credential Capture

Detects ptrace PTRACE_ATTACH/PTRACE_SEIZE syscalls recorded by auditd from 3snake, truffleproc/gdb, or any non-debugger binary attaching to three or more processes within five minutes. 3snake intercepts read/write syscalls of sshd, sudo, su, passwd and OpenLDAP to capture clear-text passwords (used by Hellhounds, Operation Lahat), and truffleproc attaches with gdb to dump process memory for secrets (T1055.008, T1003.007, T1056.001). Requires an auditd rule such as: -a always,exit -F arch=b64 -S ptrace -F a0=0x10 -k ptrace_attach.

  • Languages: Splunk SPL, Microsoft KQL, Sigma
  • ATT&CK: T1055.008, T1003.007, T1056.001
  • Data sources: linux_auditd, linux:audit, syslog
  • Severity: high
  • Confidence: medium

Splunk SPL

index=linux sourcetype="linux:audit" type=SYSCALL ((arch=c000003e syscall=101) OR (arch=c00000b7 syscall=117)) (a0=10 OR a0=4206)
| eval exe=lower(exe)
| where NOT match(exe, "/(strace|ltrace|valgrind|dlv|perf)$")
| bin _time span=5m
| stats count AS attach_calls dc(a1) AS tracee_count values(comm) AS comm values(auid) AS auid BY _time host exe
| where match(exe, "(3snake|truffleproc|gdb)") OR tracee_count>=3
| sort - tracee_count

Microsoft KQL

Syslog
| where TimeGenerated > ago(24h)
| where SyslogMessage contains "type=SYSCALL" and (SyslogMessage contains "syscall=101" or SyslogMessage contains "syscall=117")
| extend arch = extract(@'arch=([0-9a-f]+)', 1, SyslogMessage),
         syscall = extract(@'syscall=([0-9]+)', 1, SyslogMessage),
         a0 = extract(@'\ba0=([0-9a-f]+)', 1, SyslogMessage),
         tracee = extract(@'\ba1=([0-9a-f]+)', 1, SyslogMessage),
         comm = extract(@'comm="([^"]+)"', 1, SyslogMessage),
         exe = tolower(extract(@'exe="?([^\s"]+)', 1, SyslogMessage))
| where (arch =~ "c000003e" and syscall == "101") or (arch =~ "c00000b7" and syscall == "117")
| where a0 in ("10", "4206")
| where not(exe matches regex @'/(strace|ltrace|valgrind|dlv|perf)$')
| summarize AttachCalls = count(), TraceeCount = dcount(tracee), Comms = make_set(comm, 5) by Computer, exe, bin(TimeGenerated, 5m)
| where exe matches regex @'(3snake|truffleproc|gdb)$' or TraceeCount >= 3
| project TimeGenerated, Computer, exe, AttachCalls, TraceeCount, Comms

Sigma

title: Ptrace Attach by Credential Interception Tooling (3snake, truffleproc, gdb)
id: 7c1e4a52-3b9d-4f0a-9a56-2d8e1c6b7f10
status: experimental
description: Detects PTRACE_ATTACH (0x10) or PTRACE_SEIZE (0x4206) syscalls on x86_64 recorded by auditd. 3snake and truffleproc/gdb use ptrace to intercept sshd, sudo, su and passwd credentials or dump process memory. Requires an auditd rule on the ptrace syscall.
references:
  - https://dfir.ch/posts/fantastic_passwords_linux/
  - https://github.com/blendin/3snake
  - https://github.com/controlplaneio/truffleproc
  - https://global.ptsecurity.com/en/research/pt-esc-threat-intelligence/hellhounds-operation-lahat-part-2/
author: Threadlinqs Team
date: 2026-10-24
tags:
  - attack.credential-access
  - attack.defense-evasion
  - attack.privilege-escalation
  - attack.t1055.008
  - attack.t1003.007
  - attack.t1056.001
logsource:
  product: linux
  service: auditd
detection:
  selection:
    type: 'SYSCALL'
    syscall: '101'
    a0:
      - '10'
      - '4206'
  filter_debuggers:
    exe|endswith:
      - '/strace'
      - '/ltrace'
      - '/valgrind'
      - '/dlv'
      - '/perf'
  condition: selection and not filter_debuggers
fields:
  - exe
  - comm
  - a1
  - auid
falsepositives:
  - Developers or SREs debugging production processes with gdb
  - EDR, APM or profiling agents that attach with ptrace
level: high

Known false positives:

  • Administrators or developers attaching gdb to a process during legitimate troubleshooting
  • EDR, APM, profiler or crash-reporting agents that use ptrace to inspect processes

Tuning notes: Deploy the auditd ptrace rule first (-a always,exit -F arch=b64 -S ptrace -F a0=0x10 -k ptrace_attach plus a0=0x4206 for SEIZE). Whitelist approved debugger and agent binaries by exe and auid, and raise or lower the tracee_count threshold of 3 to match your environment. Set kernel.yama.ptrace_scope to 1 or higher to reduce exposure.

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)
  • Backdoored Login Portal Credential Exfiltration and Known Threat Infrastructure Contact · unlock with TL Creds
  • Linux Credential Harvesting Tool Execution (3snake, truffleproc, pspy, gdb process dump) · unlock with TL Creds
  • Shell History and Process Environment Credential Access · unlock with TL Creds
  • PAM Keystroke Audit Enablement (pam_tty_audit) for Password Capture · unlock with TL Creds
  • Yama ptrace_scope Weakened to Enable Process Interception · unlock with TL Creds
  • Entra ID Sign-in from Known Credential-Harvesting Infrastructure (UTA0178 IP) - Cloud Pivot · unlock with TL Creds
  • PAM Configuration or Module Tampering - Persistence and Credential Capture · unlock with TL Creds
  • SSH Password Logon Spread Across Multiple Linux Hosts - Lateral Movement with Harvested Credentials · unlock with TL Creds

Indicators of compromise

As of 2026-10-24, Threadlinqs Intelligence tracks 16 indicators of compromise for TL-2026-4359.

Sample indicators (3 of 16)

  • IP address: 206.189.208[.]156
  • Domain: jscloud[.]live
  • Domain: net-sensors[.]net

Defanged, TLP:CLEAR. Full set of 16 indicators: unlock with TL Creds.

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

Technical analysis

This is technique-level detection-engineering material, not a single confirmed campaign.

The DFIR.ch article 'Fantastic clear-text passwords and where to collect them (Part 1 - Linux)' (published 2026-06-22) argues that compromised Linux servers are usually pivots rather than final objectives: harvested clear-text secrets are reused for lateral movement, network pivoting and access to cloud tenants such as Azure.

Collection techniques covered: (1) Shell history review (~/.bash_history, /root/.bash_history) - the author cites a real case where a Siemens SCS Groupware keystore password sat in root's history in a keytool -importkeystore command referencing groupware.jks, alongside OpenScape XML config exposing an Azure tenant ID and client ID. (2) Process memory and environment review - /proc/<pid>/environ and truffleproc (github.com/controlplaneio/truffleproc), which attaches with GDB, dumps process memory and scans it for credential patterns. (3) pspy (github.com/DominicBreuker/pspy), which watches process command lines via inotify without root and exposes credentials passed as arguments by cron jobs and automation. (4) ptrace-based interception with 3snake (github.com/blendin/3snake), which intercepts read/write syscalls of sshd, sudo, su, ssh client, passwd and OpenLDAP and listens for process events via netlink; Positive Technologies reported the Hellhounds group using 3snake in Operation Lahat. (5) PAM abuse - native pam_tty_audit.so logging of keystrokes to audit.log, and custom PAM backdoors, often under 100 lines, that write passwords to hidden files while accepting a hardcoded credential. (6) Edge-device portal backdoors - a modified Citrix NetScaler ADC/Gateway login page posts credentials via JavaScript fetch to https://jscloud.live/items/accounts; the article references Volexity's January 2024 Ivanti Connect Secure exploitation research (CVE-2023-46805, CVE-2024-21887, UTA0178), where a modified login script exfiltrated credentials.

Defensive guidance in the source: set kernel.yama.ptrace_scope (1 restricted, 2 admin-only, 3 disabled until reboot) in /etc/sysctl.d/10-ptrace.conf; alert on PTRACE_ATTACH against sshd, sudo, su and passwd unless whitelisted; flag unrecognized binaries spawning children that attach to authentication daemons; use file integrity monitoring on /etc/pam.d/ and PAM .so files; enforce SSH key authentication and disable PasswordAuthentication in sshd_config; set HISTCONTROL=ignorespace; use a secrets manager such as HashiCorp Vault and Entra Conditional Access for workload identities.

No CVE, CVSS or per-incident exploitation confirmation is attributed to the article itself; severity is an analyst estimate.

MITRE ATT&CK techniques by tactic:

Credential Access

T1003.007 Proc Filesystem; T1056.001 Keylogging; T1056.003 Web Portal Capture; T1552.001 Credentials In Files; T1552.003 Shell History; T1556.003 Pluggable Authentication Modules

Lateral Movement

T1021.004 SSH

Defense Evasion

T1055.008 Ptrace System Calls

Privilege Escalation

T1055.008 Ptrace System Calls

Discovery

T1057 Process Discovery

Initial Access

T1078 Valid Accounts; T1078.004 Cloud Accounts; T1190 Exploit Public-Facing Application

defense-impairment

T1556.003 Pluggable Authentication Modules

Persistence

T1556.003 Pluggable Authentication Modules

Remediation

Immediate actions

  • Alert on PTRACE_ATTACH against sshd, sudo, su and passwd from non-whitelisted processes
  • Audit ~/.bash_history and /root/.bash_history for clear-text secrets and rotate any found
  • Review /etc/pam.d/ and installed PAM .so modules for unexpected or recently modified files

Workarounds

  • Set kernel.yama.ptrace_scope = 1 (or 2/3) in /etc/sysctl.d/10-ptrace.conf
  • Set HISTCONTROL=ignorespace and avoid typing secrets in commands

Longer-term hardening

  • Enforce SSH key-based authentication and set PasswordAuthentication no in /etc/ssh/sshd_config
  • Use a secrets manager (e.g. HashiCorp Vault) instead of passing secrets on command lines or in environment variables
  • Deploy file integrity monitoring on /etc/pam.d/ and PAM shared objects with high-priority alerting
  • Apply Entra Conditional Access for workload identities to limit reuse of harvested cloud credentials

Affected and fixed versions

  • Linux — Linux servers (PAM, sshd, sudo, /proc)
    Vulnerable versions: Systems without ptrace restriction, PAM integrity monitoring or secrets hygiene
  • Citrix — NetScaler ADC / Gateway authentication portal (backdoored login page example)
    Vulnerable versions: Compromised appliances

Timeline

  • Positive Technologies later reported Hellhounds had been compromising Russian organizations since at least 2021 (month approximate; year per source).
  • Earliest suspected UTA0178 activity against Ivanti Connect Secure, including credential harvesting via a modified login script (Volexity).
  • Volexity publicly disclosed exploitation of CVE-2023-46805 and CVE-2024-21887 in Ivanti Connect Secure VPN, referenced by the DFIR.ch article for portal credential interception.
  • Positive Technologies published Hellhounds: Operation Lahat Part 2, documenting 3snake use for credential harvesting on Linux (date approximate: June 2024).
  • Nextron Systems published 'Stealth in 100 Lines' analyzing PAM backdoors in Linux.
  • DFIR.ch (Stephan Berger) published 'Fantastic clear-text passwords and where to collect them (Part 1 - Linux)', covering Linux credential collection techniques and defenses.

Sources cited

Previous and next in threat intel

Drafted by the Threadlinqs research pipeline from 11 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