T1211 — Exploitation for Stealth
Tactic: Stealth · Platforms: Linux, Windows, macOS, SaaS, IaaS
Adversaries may exploit vulnerabilities to evade detection by hiding activity, suppressing logging, or operating within trusted or unmonitored components. Adversaries may exploit a system or application vulnerability to avoid detection while maintaining access within an environment. Exploitation…
Sigma detection rules (5)
T1211 Exploitation for Stealth
title: T1211 Exploitation for Stealth
id: bfbae991-55a4-47f0-935a-ad35ba851f62
description: 'Detection Strategy for Exploitation for Stealth. (Data Component: Module Load; baseline tier: windows-eventid.) Tune using: DefensiveProcessList, AllowedModulePaths, CrashThreshold.'
references:
- https://attack.mitre.org/techniques/T1211
author: Shahrukh Khan
date: 2026-01-18
license: MIT
tags:
- attack.stealth
- attack.t1211
logsource:
category: image_load
product: windows
service: sysmon
detection:
selection:
EventID: '7'
selection_target_process:
Image|endswith:
- '\MsMpEng.exe'
- '\NisSrv.exe'
- '\SenseIR.exe'
- '\SenseCncProxy.exe'
- '\SenseSampleUploader.exe'
- '\lsass.exe'
selection_unsigned:
Signed: 'false'
selection_unusual_path:
ImageLoaded|contains:
- '\Users\'
- '\ProgramData\'
- '\Temp\'
- '\AppData\'
filter_system_account:
User|contains:
- 'SYSTEM'
- 'LOCAL SERVICE'
- 'NETWORK SERVICE'
condition: selection and selection_target_process and (selection_unsigned or selection_unusual_path) and not filter_system_account
falsepositives:
- Third-party AV/EDR plugin DLLs loading into defensive processes from non-standard install paths.
level: medium
T1211 Exploitation for Stealth
title: T1211 Exploitation for Stealth
id: fe161940-5e42-483b-8822-dbd6b98245b6
description: 'Detects invocation of the systemd-coredump handler in the same event as a reference to a security-relevant process name (auditd, rsyslogd/syslog-ng/journald, sshd, fail2ban, falcon-sensor, wazuh-agentd, suricata, snort, osqueryd), which can indicate an adversary exploiting a vulnerability in one of these components to crash it and potentially suppress or evade logging/detection, per T1211. Limitation: auditd telemetry cannot itself confirm a crash was exploit-driven versus an ordinary software fault, cannot see exploitation that leaves no crash at all, and cannot observe exploitation of SaaS/cloud logging pipelines (which T1211 also covers) -- those aspects are not observable via host auditd. This replaces the prior generic any-coredump/segfault-string match, which had no logsource type and fired on routine application crashes. Tune using: WatchedBinaries, CrashPatterns.'
references:
- https://attack.mitre.org/techniques/T1211
author: Shahrukh Khan
date: 2026-02-06
license: MIT
tags:
- attack.stealth
- attack.t1211
logsource:
category: process_creation
product: linux
detection:
selection_coredump:
type: EXECVE
exe|endswith: '/systemd-coredump'
keywords_security_proc:
- 'auditd'
- 'rsyslogd'
- 'syslog-ng'
- 'journald'
- 'sshd'
- 'fail2ban'
- 'falcon-sensor'
- 'wazuh-agentd'
- 'suricata'
- 'snort'
- 'osqueryd'
condition: selection_coredump and keywords_security_proc
falsepositives:
- Security daemons can crash from resource exhaustion, kernel/library bugs, or misconfiguration unrelated to exploitation; confirming an exploit-driven crash versus an ordinary fault requires crash/core analysis auditd cannot perform.
- Package upgrades or service restarts that trigger a transient crash-then-restart cycle in a monitored daemon.
level: medium
T1211 Exploitation for Stealth
title: T1211 Exploitation for Stealth
id: 5a6d37ad-26c6-4dd1-909f-3a2ff1a781e3
description: 'Detection Strategy for Exploitation for Stealth. (Retargeted to the process_creation macOS unified-log telemetry that actually captures this techniques behavior; see logsource below.) Tune using: SecurityDaemons, UnsignedProcessThreshold.'
references:
- https://attack.mitre.org/techniques/T1211
author: Shahrukh Khan
date: 2026-02-22
license: MIT
tags:
- attack.stealth
- attack.t1211
logsource:
product: macos
detection:
selection:
Image|endswith:
- '/syspolicyd'
- '/trustd'
- '/securityd'
- '/tccd'
condition: selection
falsepositives:
- The unified log surfaces crashes and restarts of security daemons such as syspolicyd, trustd, and tccd, but distinguishing a genuine exploit-triggered crash from routine daemon restarts during updates requires binary crash-log analysis not available from this event category alone.
level: high
T1211 Exploitation for Stealth
title: T1211 Exploitation for Stealth
id: 2c5005d2-0f5c-49b6-a33a-bc0d92dce8b6
description: 'Detection Strategy for Exploitation for Stealth. (Retargeted to the cloudtrail AWS telemetry that actually captures this techniques behavior; see logsource below.) Tune using: CriticalAPIs, TimeWindow.'
references:
- https://attack.mitre.org/techniques/T1211
author: Shahrukh Khan
date: 2026-04-29
license: MIT
tags:
- attack.stealth
- attack.t1211
logsource:
category: cloudtrail
product: aws
detection:
selection:
errorCode:
- 'AccessDenied'
- 'UnauthorizedAccess'
- 'ValidationError'
selection_critical_api:
eventName:
- 'PutBucketPolicy'
- 'CreateAccessKey'
- 'AttachRolePolicy'
- 'AssumeRole'
condition: selection and selection_critical_api
falsepositives:
- Misconfigured automation and permission-boundary testing can produce the same repeated error pattern, and CloudTrail cannot distinguish a genuine exploit attempt from a benign misconfiguration without additional context, so confidence is low.
level: medium
T1211 Exploitation for Stealth
title: T1211 Exploitation for Stealth
id: ff5061b5-d251-498b-859a-193bb1f93b07
description: 'Detection Strategy for Exploitation for Stealth. (Data Component: Application Log Content; baseline tier: fallback.) Tune using: MonitoredApps, ConsentAnomalyThreshold.'
references:
- https://attack.mitre.org/techniques/T1211
author: Shahrukh Khan
date: 2026-03-02
license: MIT
tags:
- attack.stealth
- attack.t1211
logsource:
product: m365
service: unified
detection:
selection:
Operation:
- 'Consent to application.'
- 'Add OAuth2PermissionGrant.'
condition: selection
falsepositives:
- Legitimate first-time application consent grants by end users adopting new SaaS tools, since the underlying exploit mechanics of a defense evasion vulnerability are not observable in this log.
level: medium
Explore