Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
3,627 detections
Filters
Last updated
All Time
Detection languages
1,752
374
349
273
255
Contributors
908
661
593
359
164
Categories
1,489
1,153
779
528
515
Platforms
39,908
7,245
6,529
3,864
3,627
Products / Services
802
512
432
340
298
MITRE Techniques
1,179
573
523
506
438
CVEs
17
16
9
8
7
IDS Classtypes
41
15
10
6
5
IDS Protocols
47
24
3
2
1
Detects the use of 'osascript' to display credential prompts or 'dscl' with authentication-related arguments, which could indicate an attempt to phish user credentials or enumerate/create accounts on macOS systems.
This rule detects potential command and control (C2) beaconing activity by identifying devices communicating with known suspicious remote IP addresses at regular, repeating intervals. It calculates the time difference between consecutive connections to specific suspicious IPs and identifies devices that establish at least three connections within one-hour bins, adhering to a 1 to 15-minute heartbeat interval.
This rule detects network connections made to specific non-standard ports (7777, 8080, 8443) where the remote URL contains substrings associated with proxy or tunneling activities ('proxy', 'tunnel', 'hide'). This behavior is commonly associated with malware attempting to evade security controls by routing traffic through unauthorized intermediaries or obfuscated tunnels.
Detects attempts to access macOS Keychain items using 'osascript' executed from common shell processes. This activity can indicate an adversary attempting to steal credentials or other sensitive information stored in the Keychain.
Detects the use of 'rm' or 'unlink' commands to forcefully delete specific files or directories within the '/tmp' directory. The targeted files/directories ('osalogging.zip', '/tmp/sync', '/tmp/macsync_') could indicate cleanup activities after a malicious operation or system compromise, potentially related to macOS systems.
This rule detects potentially malicious activity related to the 'node-gyp' build tool used by npm. It monitors for two patterns: 1) npm spawning node-gyp, which can be an automated trigger for malicious binding.gyp files, and 2) node-gyp spawning node.exe with command-line arguments indicative of obfuscated JavaScript (e.g., base64 encoding, eval, AES-GCM patterns), a technique often seen in malicious Node.js packages.
Detects instances where the FFmpeg or FFplay multimedia frameworks spawn unexpected child processes. This behavior is indicative of successful exploitation of the RASC/DLTA codec heap overflow vulnerability, where a malicious media file triggers arbitrary code execution.
Detects instances where Claude AI-related processes (claude, claude-code) or Node.js processes associated with Claude launch suspicious commands or scripts such as setup scripts, base64-encoded commands, shell interactions, or external network connections. This behavior is indicative of potential supply chain compromise or malicious use of AI-assisted coding tools to execute arbitrary system operations.
Detects the spawning of shell or scripting processes from IDEs, code editors, or AI-assistant processes (e.g., 'claude', 'node'). The rule specifically flags potential reverse shell patterns or dangerous command-line flags often used during development or exploitation, such as '--dangerously-skip-permissions'.
This rule detects when the node.exe process initiates an outbound network connection to a specific set of non-standard, often high-numbered ports (8443, 8445, 8453, 5555, 56001-56003) over the public internet. This behavior is often indicative of potential command and control (C2) communication, reverse shells, or malicious scripts executing within a Node.js runtime environment.
Detects Python processes that combine potentially dangerous pickle deserialization (pickle.loads) with network-fetching libraries (urllib, requests, socket) and decoding methods (b64decode, fromhex). This pattern is consistent with backdoors or remote access tools (RATs) that fetch encoded payloads over the network and execute them directly in memory via deserialization.
Detects Python processes executing complex, multi-layered decoding chains (Base64, ROT13, and Hex) in their command line arguments. This pattern is often used by malicious AI skill backdoors, such as ClawHub, to obfuscate payload execution and evade signature-based detection.
Detects instances where code editors like Cursor or Claude execute shell commands or CLI tools often used for configuration, network activity, or credential interaction. This pattern may indicate malicious exploitation of editor configuration files (e.g., MCP configuration) or unauthorized use of the editor environment to execute shell commands.
This rule monitors process command line arguments for a combination of programming language interpreters (java, python, node, .NET, go) and keywords commonly associated with malware, payloads, shellcode, or injection techniques. This is intended to identify suspicious execution patterns that may indicate the staging or execution of malicious code.
This rule detects network connections or DNS queries to known public tunneling services (e.g., Cloudflare Tunnel, localtunnel, bore, devtunnels) made by processes other than common web browsers. Attackers frequently use these services to bypass firewalls and establish unauthorized reverse shells or persistent C2 channels.
Correlates a user passing a CAPTCHA gate (Cloudflare Turnstile, reCAPTCHA) with an immediate subsequent connection to an unknown, low-prevalence domain hosting the actual phishing landing page.
Identifies the macOS stealer utilizing osascript to pop a fake system dialog, requesting the user's administrator password for privilege escalation.
Detects unauthorized modification or dropping of model files directly into the local Hugging Face cache directory (~/.cache/huggingface/hub) by non-standard development processes, indicating an attempt to swap legitimate models for backdoored ones.
Detects command lines consistent with ArachneC2 implant generation and hardening options.
This rule detects pre-payload beaconing activity associated with the Meow RAT malware. It covers both network-based detection (outbound HTTP POST to known C2 domains and paths) and endpoint-based detection. Endpoint detection includes 'curl' processes communicating with known C2 infrastructure, 'osascript' processes spawned in the context of 'curl' C2 communication, and 'curl' commands using specific flags ('-k', '-s') along with a known build hash. This activity typically occurs before the main destructive payload is delivered.
Identifies the Python script explicitly reading developer secrets including AWS credentials, SSH private keys, and Kubernetes configuration files.
Page 114 of 173