Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
3,989 detections
Filters
Last updated
All Time
Detection languages
1,774
868
397
195
188
Contributors
1,070
924
756
221
96
Categories
2,448
1,307
1,164
785
713
Platforms
2,415
1,310
658
528
133
Products / Services
1,307
886
808
561
337
MITRE Techniques
18,128
15,651
12,599
8,410
6,235
3,989
CVEs
6
5
5
5
5
IDS Classtypes
564
275
11
8
6
IDS Protocols
537
174
64
64
12
Detects potentially malicious child process creation or unauthorized outbound network connections initiated by AI-assisted coding tools (e.g., Claude, Cursor, Windsurf) through their integrated Node.js or Python environments. This rule identifies common post-exploitation behaviors such as shell execution, command-line reconnaissance, and connections to external non-internal IP addresses.
Detects suspicious network traffic potentially associated with BeePRAT malware, identified by specific process image paths in temporary or TCC-related directories, connections to known malicious C2 IP addresses/ports, or non-standard TCP communication patterns involving suspicious domain queries.
This rule detects outbound network connections to a specific typosquatted domain 'perplexity-ai.online'. This domain is associated with a malicious browser extension designed to intercept and exfiltrate search queries before forwarding them to legitimate search providers.
Detects suspicious child process activity originating from common IDEs or AI language server processes that indicates AWS credential exfiltration. The rule monitors for command-line patterns indicative of credential dumping or exfiltration, such as the use of AWS CLI commands, curl, or npx with the Model Context Protocol (MCP) in conjunction with attempts to access sensitive environment variables or redirect output.
Detects network activity associated with the Djinn Stealer malware, including connections to a known C2 IP address/port, the use of a specific 'telemetry-client/1.0' user agent, or HTTP POST requests to 'application/octet-stream' endpoints originating from suspicious parent processes such as SimpleHelp or TaskWeaver.
Detects potential privilege escalation attempts against a MyBB forum by monitoring POST requests to administrative endpoints. Specifically, it flags session activity that includes creating new users with admin group assignments (usergroup=4) or accessing sensitive administrative modules (config, tools, style, user) from within the admin control panel.
Detects processes that exhibit signs of C2 communication (specifically on ports 4000-5000, associated with SystemBC) followed by the writing of executable or script files to SMB/UNC paths and subsequent remote lateral movement execution via WMI, Windows services, or scheduled tasks on the same host.
Detects rapid post-exploitation activity where a developer's workstation, upon executing VS Code (Code.exe), initiates outbound network connections to major cloud service or CI/CD provider APIs (GitHub, AWS, Azure, GCP) within 60 minutes. This behavior is indicative of the Lazarus Group 'Contagious Interview' campaign, where stolen credentials from developer environments are used to pivot into cloud infrastructure or CI/CD pipelines.
Detects web traffic containing HTTP 301/302 redirects to known cloud hosting platforms (Azure, AWS, Cloudflare) that exhibit suspicious URL patterns or originate from identified malicious referrers associated with phishing or credential harvesting campaigns.
Detects unauthorized processes reading AI assistant configuration files (e.g., Claude, Cursor, Gemini, VS Code) followed by an outbound network connection from a common scripting or runtime environment (node, python, bun) within 60 seconds, potentially indicating credential or configuration exfiltration.
Detects network connections to known IP addresses associated with FortiBleed activity, including C2, pentest labs, credential validation, and sniffer nodes. This rule helps identify potential compromise or reconnaissance related to the FortiBleed vulnerability.
Detects execution of AWS STS commands to identify the caller's identity in conjunction with suspicious outbound POST requests. This pattern is commonly used by adversaries to verify compromised cloud credentials or enumerate permissions within an AWS environment.
This rule monitors for suspicious child processes (such as bash, curl, or aws cli) spawned by the Amazon Q IDE language server. These patterns are indicative of potential abuse of IDE plugins or language servers to execute system commands, exfiltrate credentials, or interact with cloud APIs, potentially utilizing custom MCP (Model Context Protocol) configurations.
Detects anomalous activity originating from AutoGen Studio AI agents, specifically identifying attempts to abuse loopback trust boundaries via MCP (Model Context Protocol) WebSocket requests to localhost (ports 8080/8081) or the execution of suspicious child processes commonly used for command-line interference, indicative of AI agent compromise or attacker-controlled MCP command injection.
Detects the execution of a malicious file disguised as 'wusvc.exe' originating from the user Downloads folder. The rule also monitors for suspicious network connections from this process to a known C2 IP address and detects potential defensive evasion techniques such as memory protection changes (PAGE_EXECUTE_READWRITE) used for syscall stub patching or AMSI bypass in non-Microsoft signed versions of this process.
Detects non-browser processes that load 'clr.dll' (indicating in-process .NET/CLR usage) and are sending large volumes of outbound data to Google API domains (sheets.googleapis.com, oauth2.googleapis.com, googleapis.com). This behavior is consistent with the SHEETCREEP threat actor's use of Google Sheets API for command and control (C2) and data exfiltration.
Detects suspicious activity originating from web application processes (python, uvicorn, gunicorn) involving the download and execution of a payload (isp.sh), followed by the deployment and execution of a binary (lambsys) within hidden directories in /var/tmp/. This rule flags indicators of compromise such as curl/wget calls to suspicious IPs, creation of hidden directories, and execution of binary files in temporary storage locations.
Detects suspected command and control (C2) activity leveraging Cloudflare Tunnel or Workers. The rule identifies two patterns: 1) Repeated polling (>5 connections in 10 minutes) to Cloudflare infrastructure by suspicious scripting processes (e.g., PowerShell, WScript), and 2) A 'dead-drop' resolver pattern where a process contacts a Cloudflare domain, followed by a DNS query to an unrelated domain shortly thereafter, suggesting the retrieval of a new C2 address.
Detects HTTP and DNS requests directed towards the .workers.dev domain, which are commonly leveraged by threat actors for command and control (C2) redirection and domain fronting. This activity is frequently associated with the Gamaredon (Shuckworm/ACTINIUM) threat group's infrastructure.
This rule detects outbound POST requests to the Telegram Bot API (api.telegram.org) originating from internal network hosts. The detection specifically looks for requests containing the 'sendMessage' path, which is characteristic of bots transmitting stolen data (such as credentials) or command-and-control communication. It focuses on traffic initiated by common command-line tools like 'python-requests' or 'curl', which are frequently abused by automated malware scripts.
This rule detects the Python-based CastleLoader malware establishing persistent WebSocket (wss:// or ws://) command-and-control channels to randomized attacker endpoints. It specifically looks for network connections initiated by python.exe or pythonw.exe to URLs matching a WebSocket pattern, with at least 3 connections within a 5-minute window.
Page 120 of 190