Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
18,116 detections
Filters
Last updated
All Time
Detection languages
6,549
6,199
1,473
1,002
661
Contributors
3,555
2,652
2,153
2,087
1,515
Categories
7,567
3,899
1,465
1,391
1,330
Platforms
13,722
1,392
1,318
824
738
Products / Services
3,828
3,129
2,448
912
908
MITRE Techniques
18,116
15,636
12,583
8,393
6,223
CVEs
19
18
17
16
14
IDS Classtypes
305
51
24
23
22
IDS Protocols
308
136
13
10
10
Detects the use of bitsadmin.exe to download the Argamal RAT encrypted payload file (zaesdl.dat) from remote sources like GitHub. This activity is indicative of an adversary attempting to transfer malicious tools into a compromised environment.
Identifies suspicious token authorizations for the Google Chrome OAuth client. Tycoon 2FA operators frequently use this specific client and scope to proxy authentications and capture session cookies. In a complete attack sequence, this event occurs within a 1-second window alongside a login_success and a device registration event.
Detects the creation or modification of inbox rules in Exchange, which OilRig uses to hide and manage their steganographic email C2 communications.
This rule detects the use of specific command-line flags in Google Chrome processes that are associated with the FlutterShell backdoor. These flags, '--hide-crash-restore-bubble' and '--disable-session-crashed-bubble', are used by FlutterShell to suppress crash warnings after forcing a restart, indicating an attempt to evade detection or maintain persistence.
Detects the creation of the 'TieringEngineService.exe' bait file within a specific RedSun temporary directory pattern (RS-{GUID}) associated with the RedSun exploit (CVE-2026-41091). This activity indicates an attempt to exploit Windows Defender.
This rule detects a high volume of process creation events (EventID 4688) where the command line contains keywords indicative of patching or updating activity (e.g., 'patch', 'update', 'KB', 'hotfix'). Specifically, it triggers if 3 or more such events occur on a single computer within an hour. This could indicate legitimate system updates, but also potentially malicious activity masquerading as updates to evade detection or perform unauthorized actions.
Auto-Color is a Linux backdoor that installs a malicious library libcext.so.2 and writes it to /etc/ld.preload so the library loads before all other executables. This allows the malware to hook libc functions and hide its C2 connections. This rule looks for modifications to ld.preload or the presence of libcext.so.2 in the library path. In plain terms, it catches when a backdoor is trying to make itself invisible by loading first and controlling what the system reports.
Detects attempts to bypass User Account Control (UAC) by hijacking the execution flow of the auto-elevated fodhelper.exe binary via modifications to the ms-settings registry keys in the HKCU hive. Silent Stealer leverages this specific registry hijack to seamlessly gain elevated privileges without triggering user prompts, an essential step for its deep credential extraction processes.
This rule detects the download of JavaScript (.js) or HTML Application (.hta) files followed by the execution of scripting interpreters like mshta, cscript, wscript, node, or powershell with encoded commands on the same device within a one-hour window. It assigns a score based on the count of downloaded files and executed scripts, triggering an alert if the combined score exceeds 2. This behavior is indicative of potential malicious script execution, often seen in initial access or execution phases of an attack.
This rule detects suspicious DLL side-loading activity. It looks for DLL files being loaded from common user-writable or temporary directories (like `\temp\` or `\users\`) by legitimate Windows processes such as `rundll32.exe`, `explorer.exe`, or `svchost.exe`. The rule specifically flags instances where a DLL is loaded more than 3 times within an hour by these processes, indicating potential malicious activity where an adversary might be attempting to execute their own malicious DLLs by hijacking the loading process of a legitimate application.
This rule detects the execution of 'mshta.exe' when its command line contains a URL pointing to a '.7z' archive. This behavior is suspicious as mshta.exe is often abused by adversaries to download and execute malicious content, and downloading an archive directly via mshta.exe is not a typical legitimate use case.
Detects suspicious process activity originating from WSMProvHost, suggesting potential lateral movement or remote command execution
This rule detects the loading of modules (DLLs or executables) from user-specific AppData folders that are not commonly observed across the environment. It specifically excludes common legitimate paths within AppData like 'Local\Programs' and 'Local\Temp' to reduce false positives. The rule identifies modules loaded by processes that also originate from an AppData folder, indicating potentially suspicious activity such as malware or custom tools operating from user-writable locations.
Detects rundll32.exe loading content directly from a UNC path. This behavior is rare in enterprise environments and was observed in ACR Stealer delivery chains.
Detects outbound internet communications generated by rundll32.exe, which is uncommon and frequently associated with malware execution, including ACR Stealer.
Flags Intune assignments to “All Devices” or “All Users,” which attackers may abuse for mass deployment.
Detects execution of CRYPTBASE.dll from non‑system directories, a technique used in the STX RAT campaign to hijack legitimate installers (e.g., CPUID HWMonitor, X‑VPN).
Detects PowerShell commands that reverse byte arrays before dynamically loading assemblies into memory, a behavior commonly associated with advanced malware loaders and fileless execution techniques. The rule specifically looks for PowerShell process command lines containing 'IO.File', 'System.Reflection.Assembly', 'Length-1', and '..0', which are indicative of byte array manipulation and in-memory assembly loading.
Detects Hero proxyware artifacts being written or modified within the Windows SysWOW64 directory. Presence of these files may indicate installation of unauthorized proxy services or abuse of enterprise endpoints for traffic relaying.
Identifies ScreenConnect client service execution initiated by trusted system processes (dfsvc.exe, services.exe) while referencing suspicious top-level domains (.top, .info, .site, .tk, .xyz, .pw, .ml, .club, .cf, ws) in the command line. This activity may indicate unauthorized remote access deployment or abuse of legitimate remote access tools.
Detects PowerShell commands using XOR operations together with execution primitives that may indicate obfuscated in-memory payload decoding. Specifically looks for 'powershell.exe' executing with '-bxor', '81', and either 'IEX' or 'Invoke-Expression' in the command line.
Page 471 of 863