Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
714 detections
Filters
Last updated
All Time
Detection languages
330
186
84
42
40
Contributors
240
212
67
64
17
Categories
322
300
90
89
82
Platforms
585
292
181
14
11
Products / Services
10,650
9,755
6,688
4,418
3,745
714
MITRE Techniques
273
233
136
109
85
CVEs
4
4
3
3
2
IDS Classtypes
2
1
This rule detects the installation or operation of browser extensions that request overly broad host permissions (e.g., all URLs) or specifically target popular AI services (like Gemini, Copilot, or Claude). It also monitors the use of the declarativeNetRequest API to interact with these same AI service domains, which is a technique used by malicious extensions to intercept or manipulate web traffic.
This rule detects the execution of the SentinelOne memory scanner process (SentinelMemoryScanner.exe) or the loading of the associated SentinelOne core library (SentinelAgentCore.dll) from a directory path outside of the standard SentinelOne installation locations (Program Files\SentinelOne or Program Files (x86)\SentinelOne). This behavior is indicative of potential masquerading or tampering where an adversary attempts to execute a renamed or moved security tool to bypass policy or evade detection.
Detects the execution of Microsoft Teams client processes on endpoints associated with user accounts that have already been flagged as compromised via authentication anomalies. This rule acts as a pivot and investigation tool for tracking potentially malicious activity within Teams on impacted hosts.
Detects instances where PowerShell attempts to inject code into legitimate processes (such as csc.exe, chrome.exe, msedge.exe, or SearchIndexer.exe) using direct syscalls NtAllocateVirtualMemoryRemote or NtSetContextThreadRemote, indicating potential process injection activity.
This rule monitors for any process execution event captured by the EDR solution. It serves as a catch-all for recording process creation telemetry.
Detects the installation of a potentially malicious root certificate on a host. The rule monitors for common command-line arguments used by 'certutil.exe' to add certificates to the store, use of the 'security' utility for certificate installation on macOS, and the execution of processes referencing 'Localcertificate.zip'.
Detects evidence of SloppyRAT employing the Hell's Gate technique, specifically the resolution and indirect invocation of low-level NT system calls (e.g., NtAllocateVirtualMemory, NtWriteVirtualMemory, NtCreateThreadEx) designed to bypass EDR user-mode API hooking during process injection activities.
Detects the presence of the CastleLoader intermediate loader binary on disk, associated with the SloppyRAT infection chain. The rule identifies specific file content patterns, including hardcoded strings such as 'CastleLoader', 'CastleRAT', specific user-agents, and embedded Python code used for base64/zlib decoding of further payloads.
This rule is a placeholder intended to detect abuse of the OAuth device code authentication flow. The current logic attempts to identify successful logins associated with authentication failure reasons; however, the rule author notes that the necessary data source (Azure AD/Entra ID logs) is missing from the SentinelOne SDL event lake, making this detection non-functional for its stated purpose.
Detects high-frequency execution of Python interpreter processes within a short timeframe (40+ distinct command lines in 15 minutes). This behavioral pattern is indicative of automated or autonomous agent-driven payload delivery and execution, as seen in JADEPUFFER and ENCFORGE campaigns.
This rule detects the use of 'npm' or 'node' to install packages from a registry other than trusted, standard registries (e.g., npmjs.org, GitHub). It specifically monitors command lines involving 'claude-code' that utilize custom, non-standard registry URLs. This behavior is indicative of potential supply chain attacks or the installation of unauthorized/malicious dependencies.
This rule detects the use of 'npm' or 'node' to install packages from a registry other than trusted, standard registries (e.g., npmjs.org, GitHub). It specifically monitors command lines involving 'claude-code' that utilize custom, non-standard registry URLs. This behavior is indicative of potential supply chain attacks or the installation of unauthorized/malicious dependencies.
This rule detects the use of 'npm' or 'node' to install packages from a registry other than trusted, standard registries (e.g., npmjs.org, GitHub). It specifically monitors command lines involving 'claude-code' that utilize custom, non-standard registry URLs. This behavior is indicative of potential supply chain attacks or the installation of unauthorized/malicious dependencies.
This rule detects potential defense evasion activity identified by SentinelOne as call stack spoofing. The rule triggers when behavioral indicator tags related to spoofed call stacks, thread execution masking, or suspicious indicators involving RtlUserThreadStart, BaseThreadInitThunk, or kernel32.dll are detected, suggesting an attempt to hide malicious thread execution from security monitoring tools.
Detects instances where browser processes (chrome.exe or comet.exe) access multiple files in sensitive or unexpected directories (e.g., System32, /etc/, Documents) without an associated user click event. This activity often indicates potential malicious scripts or automated data collection running within the context of a browser process.
Detects instances of process call stack spoofing, specifically techniques like SparroWocky (hooking CreateThread to report fake start addresses such as AnimateWindow) or SilentMoonwalk (forged call stacks masking API call origins as RtlUserThreadStart or BaseThreadInitThunk) used to evade EDR behavioral monitoring.
Detects obfuscated PowerShell commands bypassing execution policy to download or execute 'configuration.ps1' or communicate with known malicious infrastructure associated with the ClickFix campaign. The rule specifically looks for caret-based obfuscation, bypass switches, and known IOCs in the command line.
Detects when AI-powered coding assistant plugins or CLI tools (e.g., Claude, Copilot, Gemini) execute subprocesses that attempt to access sensitive configuration files, SSH keys, or cloud credential stores. This may indicate an AI tool being manipulated or misconfigured to exfiltrate sensitive local secrets.
Detects Settra ransomware binaries characterized by naming conventions including '_win64.exe', which create 'RESTORE_FILES.txt' ransom notes and use '.locked' or '.locked_wip' file extensions for encrypted data.
This rule detects the presence of unauthorized WebView or HTML overlay windows rendered on top of active banking applications on Android devices. This behavior is indicative of credential-theft malware, such as banking trojans, which use overlays to capture user input, including banking credentials, while masking their activity over legitimate finance-related applications.
Detects the CastleRAT intermediate loader, which acts as a secondary stage before deploying SloppyRAT. The rule identifies the malicious payload by scanning for unique embedded strings, specific user-agent strings, and command-and-control (C2) communication patterns (domains and URI paths) within executable files.
Page 6 of 34