Sigma Rules
1,619 rules found
Windows Suspicious Child Process from Node.js - React2Shell
Detects suspicious child processes started by Node.js server processes on Windows, which may indicate exploitation of vulnerabilities like CVE-2025-55182 (React2Shell). Attackers can abuse the Node.js 'child_process' module to run system commands or scripts using methods such as spawn(), exec(), execFile(), fork(), or execSync(). If execSync() or exec() is used in the exploit, the command line often shows a shell (e.g., cmd.exe /d /s /c ...) running a suspicious command unless other shells are explicitly invoked. For other methods, the spawned process appears directly in the Image field unless a shell is explicitly used.
Commvault QLogin with PublicSharingUser and GUID Password (CVE-2025-57788)
Detects a qlogin.exe command attempting to authenticate as the internal `_+_PublicSharingUser_` using a GUID as the password. This could be an indicator of an attacker exploiting CVE-2025-57788 to gain initial access using leaked credentials.
Commvault QOperation Path Traversal Webshell Drop (CVE-2025-57790)
Detects the use of qoperation.exe with the -file argument to write a JSP file to the webroot, indicating a webshell drop. This is a post-authentication step corresponding to CVE-2025-57790.
Commvault QLogin Argument Injection Authentication Bypass (CVE-2025-57791)
Detects the use of argument injection in the Commvault qlogin command - potential exploitation for CVE-2025-57791. An attacker can inject the `-localadmin` parameter via the password field to bypass authentication and gain a privileged token.
Exploitation Activity of CVE-2025-59287 - WSUS Suspicious Child Process
Detects the creation of command-line interpreters (cmd.exe, powershell.exe) as child processes of Windows Server Update Services (WSUS) related process wsusservice.exe. This behavior is a key indicator of exploitation for the critical remote code execution vulnerability such as CVE-2025-59287, where attackers spawn shells to conduct reconnaissance and further post-exploitation activities.
Potential Exploitation of CVE-2025-5054 or CVE-2025-4598
Detects attempts of an attacker to enable core dumps for set-user-ID (SUID) processes by modifying the system file /proc/sys/fs/suid_dumpable, typically by setting its value to 1 or 2. Enabling this feature allows memory dumps (core dumps) of SUID processes, which usually run with elevated privileges. These dumps may contain sensitive information such as passwords, cryptographic keys or other secrets. CVE-2025-5054: Information leak via core dumps from SUID binaries using apport. CVE-2025-4598: Information disclosure in systemd-coredump due to insecure handling of SUID process memory dumps.
Atomic MacOS Stealer - FileGrabber Activity
Detects suspicious activity associated with Atomic MacOS Stealer (Amos) campaigns, including execution of FileGrabber and curl-based POST requests used for data exfiltration. The rule identifies either the execution of FileGrabber targeting /tmp or the use of curl to POST sensitive user data (including files such as /tmp/out.zip) to remote servers, which are key indicators of Amos infostealer activity.
Grixba Malware Reconnaissance Activity
Detects execution of the Grixba reconnaissance tool based on suspicious command-line parameter combinations. This tool is used by the Play ransomware group for network enumeration, data gathering, and event log clearing.
Kalambur Backdoor Curl TOR SOCKS Proxy Execution
Detects the execution of the "curl.exe" command, referencing "SOCKS" and ".onion" domains, which could be indicative of Kalambur backdoor activity.
Shai-Hulud Malware Indicators - Linux
Detects potential Shai-Hulud malware indicators based on specific command line arguments associated with its execution.
Shai-Hulud Malicious Bun Execution - Linux
Detects the execution of `bun_environment.js` via the Bun runtime, a behavior associated with the Shai-Hulud "Second Coming" NPM supply chain attack. The malware uses a `setup_bun.js` script to install the Bun runtime if not present, and then executes the malicious `bun_environment.js` payload.
Shai-Hulud 2.0 Malicious NPM Package Installation - Linux
Detects the command-line installation of specific malicious npm packages and versions associated with the Shai-Hulud 2.0 supply chain attack.
Shai-Hulud NPM Package Malicious Exfiltration via Curl
Detects potential Shai Hulud NPM package attack attempting to exfiltrate data via curl to external webhook sites.
Shai-Hulud Malware Indicators - Windows
Detects potential Shai-Hulud malware indicators based on specific command line arguments associated with its execution.
Shai-Hulud Malicious Bun Execution
Detects the execution of `bun_environment.js` via the Bun runtime, a behavior associated with the Shai-Hulud "Second Coming" NPM supply chain attack. The malware uses a `setup_bun.js` script to install the Bun runtime if not present, and then executes the malicious `bun_environment.js` payload.
Shai-Hulud 2.0 Malicious NPM Package Installation
Detects the command-line installation of specific malicious npm packages and versions associated with the Shai-Hulud 2.0 supply chain attack.
Potential CVE-2026-33829 Exploitation - Windows Snipping Tool Remote File Path URI
Detects potential exploitation of CVE-2026-33829, a vulnerability in the Windows Snipping Tool URI handler (ms-screensketch:). An attacker can abuse the 'filePath' parameter to supply a UNC path or HTTP URL, causing SnippingTool.exe to initiate a connection to a remote resource. When a UNC path is used (e.g. \\attacker.com\share), this triggers an outbound NTLM authentication attempt, allowing the attacker to capture or relay the victim's Net-NTLMv2 hash. HTTP-based paths may result in remote file loading or server-side request forgery (SSRF)-style access. The URI can be delivered via a malicious hyperlink, phishing email, or web page.
RedSun - Conhost.exe Spawned by TieringEngineService.exe
Detects two stages of the RedSun post-exploitation process chain that deliver a SYSTEM-level shell to the attacker's interactive session. Observed process chain services.exe → TieringEngineService.exe → conhost.exe (SYSTEM, CommandLine: bare path, no arguments) → cmd.exe / shell (SYSTEM, TerminalSessionId = attacker's session) Stage 1 — TieringEngineService.exe spawns argument-less conhost.exe: After winning the oplock + Cloud Files mount point race, the malicious TieringEngineService.exe (RedSun.exe copied to System32, started via CoCreateInstance / services.exe) detects it is NT AUTHORITY\SYSTEM and calls LaunchConsoleInSessionId(). This opens \\.\pipe\REDSUN, reads the attacker's session ID, duplicates the SYSTEM token, re-stamps it with that session ID via SetTokenInformation(TokenSessionId), then calls CreateProcessAsUser to spawn conhost.exe with no arguments. Stage 2 — Shell spawned from rogue conhost.exe (EDR sources with GrandParentImage): The rogue SYSTEM conhost.exe spawns a shell (cmd.exe, PowerShell, etc.) as SYSTEM in the attacker's interactive session. On EDR sources that expose GrandParentImage, the full three-level chain (TieringEngineService.exe → conhost.exe → shell) can be matched directly. The legitimate TieringEngineService.exe is a headless COM server that is unlikely to spawn conhost.exe under normal conditions.
Axios NPM Compromise Indicators - Linux
Detects the Linux-specific execution chain of the plain-crypto-js malicious npm dependency by Axios NPM package, including payload download via curl and detached execution using nohup and python3. On March 30, 2026, malicious versions (1.14.1, 0.30.4) were published to npm, injecting a dependency (plain-crypto-js@4.2.1) that executed a postinstall script as a cross-platform RAT dropper. The dropper contacted a C2 server, delivered platform-specific payloads, deleted itself, and replaced package.json to evade detection.
Axios NPM Compromise Indicators - macOS
Detects the macOS-specific execution chain of the plain-crypto-js malicious npm dependency in Axios NPM Package, including AppleScript execution via osascript, payload download, permission modification, execution, and cleanup.
Axios NPM Compromise Indicators - Windows
Detects the specific Windows execution chain and process tree associated with the Axios NPM supply chain compromise. On March 30, 2026, malicious versions (1.14.1, 0.30.4) were published to npm, injecting a dependency (plain-crypto-js@4.2.1) that executed a postinstall script as a cross-platform RAT dropper. The dropper contacted a C2 server, delivered platform-specific payloads, deleted itself, and replaced package.json to evade detection. The attack used cscript.exe (VBScript), curl.exe (C2), and PowerShell masquerading as Windows Terminal.
LiteLLM / TeamPCP Supply Chain Attack Indicators
Detects process executions related to the backdoored versions of LiteLLM (v1.82.7 or v1.82.8). In March 2026, a supply chain attack was discovered involving the popular open-source LLM framework LiteLLM by Threat Actor TeamPCP. The malicious package harvests every credential on the system, encrypts and exfiltrates them, and installs a persistent C2 backdoor.
Terminate Linux Process Via Kill
Detects usage of command line tools such as "kill", "pkill" or "killall" to terminate or signal a running process.
Process Discovery
Detects process discovery commands. Adversaries may attempt to get information about running processes on a system. Information obtained could be used to gain an understanding of common software/applications running on systems within the network
Clipboard Data Collection Via Pbpaste
Detects execution of the "pbpaste" utility, which retrieves the contents of the clipboard (a.k.a. pasteboard) and writes them to the standard output (stdout). The utility is often used for creating new files with the clipboard content or for piping clipboard contents to other commands. It can also be used in shell scripts that may require clipboard content as input. Attackers can abuse this utility in order to collect data from the user clipboard, which may contain passwords or sensitive information. Use this rule to hunt for potential abuse of the utility by looking at the parent process and any potentially suspicious command line content.
Password Protected Compressed File Extraction Via 7Zip
Detects usage of 7zip utilities (7z.exe, 7za.exe and 7zr.exe) to extract password protected zip files.
Set Files as System Files Using Attrib.EXE
Detects the execution of "attrib" with the "+s" flag to mark files as system files
Potential BOINC Software Execution (UC-Berkeley Signature)
Detects the use of software that is related to the University of California, Berkeley via metadata information. This indicates it may be related to BOINC software and can be used maliciously if unauthorized.
CMD Shell Output Redirect
Detects the use of the redirection character ">" to redirect information on the command line. This technique is sometimes used by malicious actors in order to redirect the output of reconnaissance commands such as "hostname" and "dir" to files for future exfiltration.
Potential File Override/Append Via SET Command
Detects the use of the "SET" internal command of Cmd.EXE with the /p flag followed directly by an "=" sign. Attackers used this technique along with an append redirection operator ">>" in order to update the content of a file indirectly. Ex: cmd /c >> example.txt set /p="test data". This will append "test data" to contents of "example.txt". The typical use case of the "set /p=" command is to prompt the user for input.
Headless Process Launched Via Conhost.EXE
Detects the launch of a child process via "conhost.exe" with the "--headless" flag. The "--headless" flag hides the windows from the user upon execution.
Dynamic .NET Compilation Via Csc.EXE - Hunting
Detects execution of "csc.exe" to compile .NET code. Attackers often leverage this to compile code on the fly and use it in other stages.
File Download Via Curl.EXE
Detects file download using curl.exe
Curl.EXE Execution
Detects a curl process start on Windows, which could indicates a file download from a remote location or a simple web request to a remote server
Potential Data Exfiltration Via Curl.EXE
Detects the execution of the "curl" process with "upload" flags. Which might indicate potential data exfiltration
Curl.EXE Execution With Custom UserAgent
Detects execution of curl.exe with custom useragent options
ClickOnce Deployment Execution - Dfsvc.EXE Child Process
Detects child processes of "dfsvc" which indicates a ClickOnce deployment execution.
Diskshadow Child Process Spawned
Detects any child process spawning from "Diskshadow.exe". This could be due to executing Diskshadow in interpreter mode or script mode and using the "exec" flag to launch other applications.
Diskshadow Script Mode Execution
Detects execution of "Diskshadow.exe" in script mode using the "/s" flag. Attackers often abuse "diskshadow" to execute scripts that deleted the shadow copies on the systems. Investigate the content of the scripts and its location.
Potential Proxy Execution Via Explorer.EXE From Shell Process
Detects the creation of a child "explorer.exe" process from a shell like process such as "cmd.exe" or "powershell.exe". Attackers can use "explorer.exe" for evading defense mechanisms by proxying the execution through the latter. While this is often a legitimate action, this rule can be use to hunt for anomalies. Muddy Waters threat actor was seeing using this technique.
Potential DLL Sideloading Activity Via ExtExport.EXE
Detects the execution of "Extexport.exe".A utility that is part of the Internet Explorer browser and is used to export and import various settings and data, particularly when switching between Internet Explorer and other web browsers like Firefox. It allows users to transfer bookmarks, browsing history, and other preferences from Internet Explorer to Firefox or vice versa. It can be abused as a tool to side load any DLL. If a folder is provided in the command line it'll load any DLL with one of the following names "mozcrt19.dll", "mozsqlite3.dll", or "sqlite.dll". Arbitrary DLLs can also be loaded if a specific number of flags was provided.
Potential Password Reconnaissance Via Findstr.EXE
Detects command line usage of "findstr" to search for the "passwords" keyword in a variety of different languages
New Self Extracting Package Created Via IExpress.EXE
Detects the "iexpress.exe" utility creating self-extracting packages. Attackers where seen leveraging "iexpress" to compile packages on the fly via ".sed" files. Investigate the command line options provided to "iexpress" and in case of a ".sed" file, check the contents and legitimacy of it.
Microsoft Workflow Compiler Execution
Detects the execution of Microsoft Workflow Compiler, which may permit the execution of arbitrary unsigned code.
CodePage Modification Via MODE.COM
Detects a CodePage modification using the "mode.com" utility. This behavior has been used by threat actors behind Dharma ransomware.
Net.EXE Execution
Detects execution of "Net.EXE".
SMB over QUIC Via Net.EXE
Detects the mounting of Windows SMB shares over QUIC, which can be an unexpected event in some enterprise environments.
Suspicious New Instance Of An Office COM Object
Detects an svchost process spawning an instance of an office application. This happens when the initial word application creates an instance of one of the Office COM objects such as 'Word.Application', 'Excel.Application', etc. This can be used by malicious actors to create malicious Office documents with macros on the fly. (See vba2clr project in the references)