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 ClickFix-style browser or explorer spawned command execution using clipboard-paste patterns, or subsequent ransomware-related discovery and inhibit-system-recovery commands such as net.exe, nltest.exe, and vssadmin.exe used after initial access.
Detects a multi-stage Telephone-Oriented Attack Delivery (TOAD) campaign. The first stage identifies phishing emails containing billing or security-themed subjects and phone numbers without links or attachments to evade automated analysis. The second stage monitors for the subsequent installation or execution of common remote-support tools (e.g., AnyDesk, TeamViewer, ScreenConnect). A final correlative rule detects these events occurring sequentially on the same host, indicating a high-confidence indicator of a successful callback phishing social-engineering attempt.
Detects a two-stage activity: first, the potential unauthorized reading or access of LLM provider API keys from files, environment variables, or command-line arguments; and second, the subsequent use of those same credentials to authenticate to LLM provider APIs (OpenAI, Anthropic, Azure, Bedrock) from a different, unrecognized source IP address, indicating potential credential theft and session hijacking.
Detects a two-stage activity: first, the potential unauthorized reading or access of LLM provider API keys from files, environment variables, or command-line arguments; and second, the subsequent use of those same credentials to authenticate to LLM provider APIs (OpenAI, Anthropic, Azure, Bedrock) from a different, unrecognized source IP address, indicating potential credential theft and session hijacking.
Detects anomalous, chained execution patterns initiated by AI agents integrated with Model Context Protocol (MCP) servers. The rule identifies a multi-stage sequence involving the invocation of browser automation tools followed by shell execution and subsequent outbound network activity, potentially indicating an AI-driven attack chaining filesystem, shell, or cloud-API access in an unauthorized manner.
Detects anomalous, rapid execution of a high volume of diverse offensive security tools (scanners, exploitation frameworks, and C2 agents) from a single parent process within a short window, which is characteristic of automated LLM-driven orchestration or scripted attack tool chaining.
Detects endpoint behavior consistent with AI-assisted exploit development, characterized by the execution of reverse engineering and vulnerability research tools (e.g., IDA Pro, Ghidra, AFL) interleaved with frequent outbound network connections to LLM APIs, followed by the local compilation of a new executable or DLL.
Detects unauthorized modifications to AI agent configuration files, including system prompts, tool allow-lists, and MCP server registration files. Such changes are often performed by processes outside of approved configuration management or infrastructure-as-code pipelines, potentially allowing for persistent manipulation of AI model behavior and capabilities.
Detects anomalous, chained execution patterns initiated by AI agents integrated with Model Context Protocol (MCP) servers. The rule identifies a multi-stage sequence involving the invocation of browser automation tools followed by shell execution and subsequent outbound network activity, potentially indicating an AI-driven attack chaining filesystem, shell, or cloud-API access in an unauthorized manner.
Detects anomalous, chained execution patterns initiated by AI agents integrated with Model Context Protocol (MCP) servers. The rule identifies a multi-stage sequence involving the invocation of browser automation tools followed by shell execution and subsequent outbound network activity, potentially indicating an AI-driven attack chaining filesystem, shell, or cloud-API access in an unauthorized manner.
Detects anomalous, rapid execution of a high volume of diverse offensive security tools (scanners, exploitation frameworks, and C2 agents) from a single parent process within a short window, which is characteristic of automated LLM-driven orchestration or scripted attack tool chaining.
Detects outbound HTTPS connections to major generative AI inference endpoints originating from non-browser, non-development tool processes at regular intervals. This activity is indicative of beaconing behavior, where malware (specifically families like LAMEHUG) leverages LLM APIs for dynamic command generation, payload obfuscation, or C2 instruction retrieval, bypassing traditional security controls.
Detects endpoint behavior consistent with AI-assisted exploit development, characterized by the execution of reverse engineering and vulnerability research tools (e.g., IDA Pro, Ghidra, AFL) interleaved with frequent outbound network connections to LLM APIs, followed by the local compilation of a new executable or DLL.
Detects anomalous, high-volume file access by a single process, followed by local file compression or archiving, and subsequent outbound data transfer to a non-internal IP address. This pattern is characteristic of automated agents (like those used in the GTG-1002/Anthropic AI-orchestrated campaigns) performing rapid data triage and staging for exfiltration.
Detects anomalous child process spawning from common web browsers or productivity applications (e.g., Office, Acrobat) on hosts that have recently communicated with known LLM or code-generation APIs. This behavioral correlation is intended to identify the potential execution of AI-generated exploit code.
Detects the execution of common remote monitoring and management (RMM) software on a host shortly after the associated user account underwent a password or MFA reset, a pattern indicative of help-desk impersonation (vishing) followed by unauthorized remote access setup.
Detects instances where AI-assisted development tools (such as Claude, Codex, Copilot, or Gemini) initiate command-line processes (e.g., shells, interpreters, or network utilities) with arguments indicative of credential access, reconnaissance, or sensitive file interactions.
This rule detects when the Node.js runtime process (node.exe) initiates the Windows command interpreter (cmd.exe) with the '/c' flag, which is commonly used to execute commands or scripts. This behavior is frequently associated with malicious Node.js activity, such as Remote Access Trojans (RATs) or web shells, which may use the shell to execute system-level operations or bypass typical application constraints.
Detects executable files launched from torrent-related directories or parent processes that perform reconnaissance commands or registry queries indicative of sandbox or virtualization detection, a technique observed in the MovieReaper malware loader.
Detects network connections to known SideCopy/ReverseRAT command-and-control infrastructure, including a static C2 IP address and two C2/hosting domains (matched exactly and as proper subdomains to avoid substring false positives).
Detects a suspicious sequence of events where a process creates a scheduled task named 'WindowsUpdate' followed by execution of 'netsh' commands to dump wireless profiles or clear keys within a 15-minute window. This behavioral pattern is often associated with credential-stealing malware attempting to establish persistence and harvest sensitive information.
Page 63 of 465