Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
988 detections
Filters
Last updated
All Time
Detection languages
477
129
119
73
69
Contributors
362
161
155
68
43
Categories
20,456
11,839
5,876
5,155
4,865
988
Platforms
568
341
38
36
22
Products / Services
234
206
170
158
109
MITRE Techniques
560
464
381
238
192
CVEs
23
23
15
13
12
IDS Classtypes
1
IDS Protocols
1
Detects attempts to modify internal PostgreSQL system catalog tables (pg_authid or pg_proc) to grant superuser privileges or alter authorization records. This activity is commonly associated with exploiting logical decoding plugins to achieve persistence or privilege escalation in PostgreSQL environments.
Detects direct SQL UPDATE operations on the pg_authid system table to modify privilege flags (e.g., rolsuper, rolcreaterole) without using the authorized ALTER ROLE statement. Such activity often indicates an attempt to escalate privileges or modify database user permissions in an unauthorized manner, bypassing standard auditing and role management processes.
Detects token duplication/impersonation targeting winlogon.exe to obtain SYSTEM privileges (SeDebugPrivilege abuse, DuplicateToken, SetTokenInformation to Untrusted integrity)
This rule detects the loading of specific, known-vulnerable kernel drivers (based on their SHA256 hashes) when triggered by administrative command-line utilities such as cmd.exe or powershell.exe. This behavior is highly characteristic of 'Bring Your Own Vulnerable Driver' (BYOVD) attacks, where adversaries exploit vulnerabilities in legitimate, signed drivers to gain elevated (often SYSTEM or kernel) privileges.
This rule detects the termination of critical security software processes (e.g., antivirus agents) by a process that has recently acquired SeDebugPrivilege. This behavior is indicative of a malicious actor attempting to disable endpoint security controls to evade detection.
This rule detects the termination of critical security software processes (e.g., antivirus agents) by a process that has recently acquired SeDebugPrivilege. This behavior is indicative of a malicious actor attempting to disable endpoint security controls to evade detection.
This rule detects potential tampering with the Huorong security product by monitoring for specific process names (HipsTray.exe, HipsMain.exe, HipsDaemon.exe, wsctrlsvc.exe, TrafficProt.exe) interacting with suspicious API calls or command lines indicative of token privilege manipulation, service blinding, or security product disabling/downgrade.
This rule detects attempts to manipulate process access tokens using the Windows 'SetTokenInformation' API. Specifically, it monitors for adjustments to token privileges or integrity levels, or the presence of 'Untrusted' markers. These behaviors are often indicative of an adversary attempting to bypass Windows access controls, perform token manipulation, or conduct privilege escalation.
Detects the loading of potential Bring Your Own Vulnerable Driver (BYOVD) drivers, specifically those known to be utilized for bypassing endpoint security controls. This rule triggers on specific driver filenames or drivers loaded from non-standard, user-writable directories (e.g., %TEMP%, Public), which are common tactics for dropping and executing malicious kernel-mode code.
Detects instances where the same user performs both a role assignment (potentially for privilege escalation) and executes a Run Command on an Azure Virtual Machine within a one-hour window. This behavior is indicative of an attacker modifying access control and immediately executing arbitrary code on the target resource.
This rule performs a two-layer hunt for the ShieldBreak/RoguePlanet (CVE-2026-50656) exploit. Layer 1 detects potential privilege escalation by identifying Windows Error Reporting processes (WerFault, WerFaultSecure, or WerMgr) running as SYSTEM that initiate suspicious child processes like cmd.exe, powershell.exe, or rundll32.exe. Layer 2 identifies the presence of known exploit artifacts including 'ShieldBreak.exe', 'Warden.dll', and the 'eicar_com.zip' test file via process execution or file creation events.
This rule performs a two-layer hunt for the ShieldBreak/RoguePlanet (CVE-2026-50656) exploit. Layer 1 detects potential privilege escalation by identifying Windows Error Reporting processes (WerFault, WerFaultSecure, or WerMgr) running as SYSTEM that initiate suspicious child processes like cmd.exe, powershell.exe, or rundll32.exe. Layer 2 identifies the presence of known exploit artifacts including 'ShieldBreak.exe', 'Warden.dll', and the 'eicar_com.zip' test file via process execution or file creation events.
Detects instances where a non-SYSTEM process attempts to open a handle to winlogon.exe. Since winlogon.exe typically runs as SYSTEM and manages user sessions, unauthorized access to its process handle is often a precursor to credential dumping or token manipulation attacks aimed at escalating privileges to SYSTEM.
This rule detects attempts to manipulate process access tokens using the Windows 'SetTokenInformation' API. Specifically, it monitors for adjustments to token privileges or integrity levels, or the presence of 'Untrusted' markers. These behaviors are often indicative of an adversary attempting to bypass Windows access controls, perform token manipulation, or conduct privilege escalation.
Detects instances where the Windows Defender antimalware service (MsMpEng.exe) spawns common shell processes (cmd.exe, powershell.exe, pwsh.exe, or conhost.exe) while running under the SYSTEM account. This behavior is highly anomalous and indicative of potential exploitation involving privilege escalation via the Microsoft Defender process.
The following analytic detects WerMgr.exe (Windows Error Reporting) spawning a child process running at SYSTEM integrity level.
WerMgr.exe normally runs at the integrity level of the reporting user or as a background SYSTEM-owned service that does not launch interactive children.
In the ShieldBreak exploit, WerMgr.exe is manually triggered via the QueueReporting scheduled task and loads an attacker-planted phantom DLL (phoneinfo.dll), which then spawns an elevated shell.
If confirmed malicious, this activity indicates successful local privilege escalation to SYSTEM.
WerMgr.exe normally runs at the integrity level of the reporting user or as a background SYSTEM-owned service that does not launch interactive children.
In the ShieldBreak exploit, WerMgr.exe is manually triggered via the QueueReporting scheduled task and loads an attacker-planted phantom DLL (phoneinfo.dll), which then spawns an elevated shell.
If confirmed malicious, this activity indicates successful local privilege escalation to SYSTEM.
The following analytic detects mpclient.dll, the Windows Defender client API library, being loaded by a process that is not part of the Windows Defender platform.
mpclient.dll exposes the API surface used to drive on-demand Defender scans (IOAV, MpScan). In the ShieldBreak exploit, the attacker binary loads mpclient.dll directly and calls its scan APIs against an object-manager path in order to trigger a Defender scan against attacker-controlled content as part of a race condition targeting Defender's placeholder-hydration behavior.
If confirmed malicious, this activity indicates an attempt to weaponize Windows Defender's own scanning pipeline for local privilege escalation.
mpclient.dll exposes the API surface used to drive on-demand Defender scans (IOAV, MpScan). In the ShieldBreak exploit, the attacker binary loads mpclient.dll directly and calls its scan APIs against an object-manager path in order to trigger a Defender scan against attacker-controlled content as part of a race condition targeting Defender's placeholder-hydration behavior.
If confirmed malicious, this activity indicates an attempt to weaponize Windows Defender's own scanning pipeline for local privilege escalation.
Detects SYSTEM processes initiated by non-service user accounts from user-writable locations such as AppData, Temp, Public, ProgramData, and Downloads. This behavior may indicate local privilege escalation and is relevant to post-exploitation activity associated with CVE-2026-68820.
The following analytic detects an unprivileged user invoking the unshare syscall with user namespace flags followed within 120 seconds by a root-owned shell or interpreter spawning under the same parent process, correlating auditd syscall telemetry with Sysmon process creation events to identify the two-step sequence characteristic of user-namespace-based Linux kernel privilege escalation exploits such as DirtyFrag.
The following analytic detects when the AF_ALG kernel crypto socket interface being loaded more than 300 seconds after system boot, which is a primary kernel-level indicator of Copy Fail (CVE-2026-31431) exploitation activity on Debian and Ubuntu family systems.
The AF_ALG interface is required by the exploit to access the vulnerable authencesn crypto code path, and on systems where it is not auto-loaded at boot, its on-demand registration by an unprivileged process is a strong indicator of exploitation in progress.
The AF_ALG interface is required by the exploit to access the vulnerable authencesn crypto code path, and on systems where it is not auto-loaded at boot, its on-demand registration by an unprivileged process is a strong indicator of exploitation in progress.
Linux Unexpected Process Spawning su
Cortex XDR
Detects instances where a non-root user process launches 'su' or a shell command executing 'su' from an unusual parent process (excluding standard shells like bash/sh/zsh/ksh or elevation tools like sudo/su). This is used to identify suspicious attempts to switch user context or escalate privileges.
Page 18 of 48