Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
4,464 detections
Filters
Last updated
All Time
Detection languages
1,883
1,282
329
251
239
Contributors
1,186
586
532
507
347
Categories
2,846
2,669
2,391
1,718
312
Platforms
4,409
153
89
34
21
Products / Services
1,182
1,146
1,042
668
277
MITRE Techniques
18,119
15,639
12,589
8,399
6,227
4,464
CVEs
12
6
6
6
6
IDS Classtypes
20
7
2
1
1
IDS Protocols
22
5
3
1
Detects PowerShell execution using Get-CimInstance to query Win32_ComputerSystem, Win32_BIOS, Win32_BaseBoard, or other WMI classes commonly checked by malware to detect virtual machine or sandbox environments. This technique is frequently used by malware families to evade analysis in virtualized security research environments.
This rule detects suspicious command-line activity involving 'conhost.exe' running in headless mode, combined with specific keywords often associated with malware or malicious scripts. The presence of '--headless' with other suspicious strings like 'svchost.exe', 'certutil -decode', common temporary directories, or specific file extensions/initiating processes (cmd.exe, powershell.exe, curl) suggests an attempt to execute hidden or obfuscated malicious code. This pattern is frequently observed in various stages of an attack, including execution, persistence, and defense evasion.
Power query to find Base64 encoded commands
Detects the execution of Node.js (node.exe) when spawned by PowerShell or the Command Prompt from a user-writable directory. This behavior is characteristic of the Tsundere Bot delivery chain used by threat actors like TA584 (Storm-0900), who use a PowerShell script to install and run the Node.js-based malware.
Detects the SmarterMail service process ('MailService.exe') spawning command-line interpreters to execute reconnaissance commands. This behavior is highly indicative of remote code execution following the exploitation of vulnerabilities like CVE-2026-23760.
Detects when a Python process spawns a shell or a network utility. This behavior can be indicative of Remote Code Execution (RCE) vulnerabilities being exploited in AI/ML libraries (e.g., CVE-2025-23304, CVE-2026-22584) that allow attackers to execute arbitrary code when a malicious model file is loaded.
Detects the use of a command injection vulnerability in the legitimate Windows script 'SyncAppvPublishingServer.vbs' to execute arbitrary code. As seen in the ClearFake campaign, attackers can append a semicolon to the command-line arguments to execute subsequent malicious PowerShell commands.
Detects the execution of a PowerShell command that reads a file's byte stream and performs a XOR operation on it. This technique is used by the KONNI APT group to decrypt and execute an embedded payload in memory, as seen in their 2026 campaign involving AI-generated backdoors.
This rule detects attempts to establish persistence by modifying the 'RunMRU' registry key to execute malicious commands. It specifically looks for the presence of common scripting and download utilities (powershell, pwsh, mshta, certutil, bitsadmin, curl, wget) combined with indicators of remote execution or code download (http, iex, Invoke-Expression, DownloadString, iwr, Invoke-WebRequest) within the registry data.
This rule detects attempts to use the Windows registry key 'Software\Microsoft\Windows\CurrentVersion\Explorer\TypedPaths' to execute potentially malicious commands. It specifically looks for the presence of 'powershell', 'mshta', 'certutil', 'curl', 'bitsadmin', or 'wget' within the registry data associated with this key. This technique can be used for persistence or execution of malicious payloads.
Detects Remcos RAT payloads and commands
Detects obfuscated PowerShell commands used to disable security features, such as AMSI bypass, disabling TLS certificate validation, and enabling Restricted Admin mode. This behavior is a known TTP of the Qilin ransomware group.
Detects the execution of PowerShell with command-line arguments containing keywords indicative of in-memory download and execution, such as '-enc' or '-EncodedCommand' used in conjunction with 'IEX' (Invoke-Expression) or 'DownloadString'. This pattern is a common TTP for initial access and second-stage payload delivery, and was observed in the exploitation of CVE-2025-55182 (React2Shell).
Detects a PowerShell process that makes an outbound network connection and contains command-line arguments associated with reflectively loading a .NET assembly from a Base64 encoded string. This is a common fileless technique used by malware like Remcos to execute payloads in memory.
Detects the SHADOW#REACTOR process execution chain of wscript.exe spawning powershell.exe, which in turn spawns MSBuild.exe.
Detects PowerShell commands that utilize string normalization obfuscation techniques before Base64 decoding. This rule identifies a generic pattern where one or two placeholder characters are replaced with a literal character, a technique used by malware like SHADOW#REACTOR to bypass static string analysis of the encoded payload.
Detects the use of PowerShell to download and/or expand a Python environment from the official python.org website. This technique has been observed in AsyncRAT campaigns to stage a legitimate interpreter for executing malicious Python code, bypassing defenses that might block known malicious executables.
Detects when a Python process spawns a shell or a network utility. This behavior can be indicative of Remote Code Execution (RCE) vulnerabilities being exploited in AI/ML libraries (e.g., CVE-2025-23304, CVE-2026-22584) that allow attackers to execute arbitrary code when a malicious model file is loaded.
Detects suspicious commands written to the Windows RunMRU registry key, which stores the history of commands executed via the Windows Run dialog (Win+R). This activity is often associated with the "ClickFix" social engineering technique, where users are tricked into pasting and running malicious commands. The rule identifies various suspicious indicators, including fake verification messages, PowerShell commands with obfuscation or download cradles, mshta, wscript, curl, wget, rundll32, bitsadmin, ssh proxy abuse, and obfuscated command line arguments.
Detects a suspicious process chain involving wscript.exe spawning powershell.exe, which then spawns MSBuild.exe. This specific chain is characteristic of the SHADOW#REACTOR malware campaign, indicating potential execution of malicious code via trusted utilities.
Detects a multi-stage process execution chain characteristic of the 'ClickFix' technique. This involves 'explorer.exe' spawning a headless 'conhost.exe', which subsequently executes a Living Off the Land Binary (LOLBin) such as 'curl', 'pwsh', 'powershell', or 'mshta' to download a payload from an HTTP/HTTPS source. This chain indicates an attempt to bypass traditional security controls and execute malicious code.
Page 154 of 213