Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
9,759 detections
Filters
Last updated
All Time
Detection languages
3,724
2,036
1,335
1,055
969
Contributors
2,963
2,050
1,372
1,226
485
Categories
5,718
3,075
1,138
900
659
Platforms
9,646
658
432
53
30
Products / Services
10,696
9,759
6,696
4,421
3,752
MITRE Techniques
3,828
3,376
1,980
1,617
1,042
CVEs
18
14
12
12
11
Detects anomalous behavioral patterns where an AI agent process (e.g., Claude Code, Cursor, Aider) performs multi-category discovery of its own runtime environment, plugins, and host configuration. The rule identifies agents that trigger processes spanning at least three distinct discovery categories (Network, File/Plugin, Software, or System) within a short timeframe, indicating potential reconnaissance of agent-authorized capabilities and tools rather than standard host exploration. Covers T1082, T1518, T1083, T1016
Detects anomalous behavioral patterns where an AI agent process (e.g., Claude Code, Cursor, Aider) performs multi-category discovery of its own runtime environment, plugins, and host configuration. The rule identifies agents that trigger processes spanning at least three distinct discovery categories (Network, File/Plugin, Software, or System) within a short timeframe, indicating potential reconnaissance of agent-authorized capabilities and tools rather than standard host exploration. Covers T1082, T1518, T1083, T1016
Detects a multi-stage, automated attack chain performed by AI-based development agents (e.g., Claude, Copilot, Cursor). The rule correlates process discovery, credential access, file collection (staging), and exfiltration over web services within a tight 15-minute window, identifying potential autonomous exploitation and data exfiltration. Covered T1046, T1552.001, T1078, T1005, T1567
Detects a multi-stage, automated attack chain performed by AI-based development agents (e.g., Claude, Copilot, Cursor). The rule correlates process discovery, credential access, file collection (staging), and exfiltration over web services within a tight 15-minute window, identifying potential autonomous exploitation and data exfiltration. Covered T1046, T1552.001, T1078, T1005, T1567
Detects a multi-stage, automated attack chain performed by AI-based development agents (e.g., Claude, Copilot, Cursor). The rule correlates process discovery, credential access, file collection (staging), and exfiltration over web services within a tight 15-minute window, identifying potential autonomous exploitation and data exfiltration. Covered T1046, T1552.001, T1078, T1005, T1567
Detects the WAV-based loader staging technique: a RIFF/WAVE file carrying a high-entropy appended blob, or a companion DLL chain (rdCore.dll/WMPCL.dll/WPFLocalizeExtension.dll) referencing the hidden audio payload and the encrypted monitor.raw container used to unpack the final RAT
Detects the WAV-based loader staging technique: a RIFF/WAVE file carrying a high-entropy appended blob, or a companion DLL chain (rdCore.dll/WMPCL.dll/WPFLocalizeExtension.dll) referencing the hidden audio payload and the encrypted monitor.raw container used to unpack the final RAT
Detects the WAV-based loader staging technique: a RIFF/WAVE file carrying a high-entropy appended blob, or a companion DLL chain (rdCore.dll/WMPCL.dll/WPFLocalizeExtension.dll) referencing the hidden audio payload and the encrypted monitor.raw container used to unpack the final RAT
This rule detects scenarios where an AI agent executes shell commands or system-level tools immediately after receiving a prompt injection attack from an untrusted source. It correlates agent audit logs—specifically identifying tool calls flagged for prompt injection—with actual endpoint process creation events. It flags process execution as 'High' severity if the command line matches common malicious patterns like encoded commands, IEX, or attempts to access credential material (e.g., lsass, mimikatz).
This rule detects scenarios where an AI agent executes shell commands or system-level tools immediately after receiving a prompt injection attack from an untrusted source. It correlates agent audit logs—specifically identifying tool calls flagged for prompt injection—with actual endpoint process creation events. It flags process execution as 'High' severity if the command line matches common malicious patterns like encoded commands, IEX, or attempts to access credential material (e.g., lsass, mimikatz).
Detects the creation of a new local account that is immediately followed by (or associated with) execution from a process identified as an AI development or agent-based tool (e.g., Cursor, GitHub Copilot, Ollama). The rule correlates process creation events with Windows Security Event ID 4720 (User account created) and optionally checks for subsequent addition to the local Administrators group. This logic aims to surface potential unauthorized credential creation triggered by automated AI coding assistants or LLM-driven workflows.
Detects the creation of a new local account that is immediately followed by (or associated with) execution from a process identified as an AI development or agent-based tool (e.g., Cursor, GitHub Copilot, Ollama). The rule correlates process creation events with Windows Security Event ID 4720 (User account created) and optionally checks for subsequent addition to the local Administrators group. This logic aims to surface potential unauthorized credential creation triggered by automated AI coding assistants or LLM-driven workflows.
Detects the creation of a new local account that is immediately followed by (or associated with) execution from a process identified as an AI development or agent-based tool (e.g., Cursor, GitHub Copilot, Ollama). The rule correlates process creation events with Windows Security Event ID 4720 (User account created) and optionally checks for subsequent addition to the local Administrators group. This logic aims to surface potential unauthorized credential creation triggered by automated AI coding assistants or LLM-driven workflows.
Detects the creation of a new local account that is immediately followed by (or associated with) execution from a process identified as an AI development or agent-based tool (e.g., Cursor, GitHub Copilot, Ollama). The rule correlates process creation events with Windows Security Event ID 4720 (User account created) and optionally checks for subsequent addition to the local Administrators group. This logic aims to surface potential unauthorized credential creation triggered by automated AI coding assistants or LLM-driven workflows.
Detects service-creation commands (sc create / New-Service / reg add under a Services key) whose COMMAND LINE itself references a known NeedyMantis binary name or sideload staging folder. Requiring the malicious reference to appear in the command being executed -- rather than merely in the calling process's own folder location -- removes false matches from unrelated legitimate software that happens to be installed under a similarly-named folder.
Detects the execution of PowerShell or PowerShell ISE where the parent process is the console host (conhost.exe) and the command-line arguments are either missing, empty, or represent an bare execution of the binary. This pattern is commonly associated with fileless execution techniques where attackers attempt to hide command-line arguments or evade logging.
Detects the deployment of NetSupport Manager (client32.exe) or related components when initiated by common scripting interpreters such as PowerShell, MSHTA, or CMD. This behavior is indicative of a ClickFix-style social engineering attack where a user is coerced into executing malicious commands to deploy remote access tools.
Detects instances where cmd.exe or powershell.exe are launched from explorer.exe with command-line arguments containing DocuSign-themed keywords. This behavior is indicative of a 'ClickFix' phishing attack, where a user is socially engineered to copy and execute a malicious command via the terminal after interacting with a fraudulent DocuSign lookalike page.
Detects the silent installation or execution of Remote Monitoring and Management (RMM) tools when launched by script hosts (wscript.exe, mshta.exe, powershell.exe) which themselves were spawned by browser processes or explorer.exe. This pattern is characteristic of social engineering delivery chains, such as 'ClickFix', where a user is tricked into executing a script that subsequently fetches and runs RMM payloads for persistent remote access.
Detects the execution of the Chisel C2 tool by matching its unique command-line grammar, specifically the 'client' subcommand combined with reverse remote ('R:') configurations. This detection strategy is resilient against binary renaming and infrastructure rotation by focusing on the tool's required argument structure rather than static IOCs like filenames or C2 addresses. It is specifically useful for identifying unauthorized remote tunnels used for persistence and data exfiltration.
Detects instances where a Claude process accesses a 'SKILL.md' file, followed closely by a suspicious command execution on the same device. The rule specifically targets command-line activity that involves encoding, download-and-execute patterns, staging in sensitive directories, or communication with suspicious domains/IPs, which is indicative of a supply-chain or poisoned agent-skill file attack.
Page 61 of 465