Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
2,686 detections
Filters
Last updated
All Time
Detection languages
1,342
389
264
219
197
Contributors
620
506
358
209
102
Categories
20,456
11,839
5,876
5,155
4,865
2,686
Platforms
1,168
638
452
382
359
Products / Services
593
445
403
264
229
MITRE Techniques
1,884
787
525
521
510
CVEs
11
5
5
5
3
IDS Classtypes
189
133
34
10
7
IDS Protocols
243
58
33
22
12
Detects a non-browser mailbox collection or archive utility connecting directly to an IP address while RemoteUrl is empty. This behavior is suspicious as legitimate email clients typically use hostnames. The rule filters out common browsers and svchost.exe and looks for keywords in the command line indicative of mailbox or archive activity (e.g., .pst, .ost, Aspose.Email, mailbox, archive). This suggests an attempt to exfiltrate or collect email data, potentially bypassing DNS resolution for C2 or evasion.
Detects HTTP network connections to the specific URI path and parameters used by the OnyxC2 stealer for C2 beaconing and data exfiltration.
Detects HTTPS POST requests to known attacker-controlled infrastructure associated with the BadBlocker campaign (e.g., Unistream SDK, adblock-for-youtube domains). These requests often originate from browser extensions via injected scripts, facilitating the potential exfiltration of sensitive browser session data or credentials.
Detects behavior indicative of the 'Adblock for YouTube' browser extension (or associated malicious variants) fetching rule configuration files from 'api.adblock-for-youtube.com', followed by suspicious outbound network traffic to non-Google/YouTube domains. This pattern suggests the extension is performing unauthorized data exfiltration, such as harvesting session cookies or credentials via injected scripts.
Detects credential and session data exfiltration by browsers communicating with Bluekit PhaaS-associated domains or Telegram API webhooks. The rule filters network traffic originating from common web browsers to known malicious domains associated with Bluekit infrastructure or to the public Telegram API for automated exfiltration.
Detects the execution of a malicious npm package (scope squatting on @withgoogle/stitch-sdk) that runs a preinstall script designed to harvest sensitive credentials (Claude, Git, SSH, GitHub, npm, Docker) and exfiltrate them to the domain 'stitch-production.org'.
Detects explorer.exe making DNS requests to Discord or Telegram, which is anomalous and indicates potential process injection for data exfiltration. This behavior is associated with information stealer malware like Phantom Stealer.
Detects administrative mail, forwarding, or content compliance rule changes performed from an IP address classified as an anonymous proxy. This behavior is often associated with threat actors like UNC6508, who use obfuscation networks and proxies to access administrator accounts and exfiltrate data.
Detects the infostealer communicating with raw.githubusercontent.com to pull secondary modules or exfiltrate small data fragments via URL params, specifically when initiated by 'python.exe'.
Identifies scripts or commands iterating through logical drives specifically searching for sensitive file extensions (.kdbx, .pdf, .docx, .xlsx, .txt) and then creating a text file, indicating potential data staging prior to exfiltration.
Detects processes containing sensitive credential-related keywords (cookies, creds, password, token) in their command lines followed by a network connection to known suspicious remote IP addresses within a 60-minute window, suggesting potential credential exfiltration.
Detects processes attempting to exfiltrate data over a Telegram C2 channel by making network connections to api.telegram.org or t.me, excluding legitimate Telegram client binaries and common web browsers. This behavior is characteristic of Vidar Stealer and other Malware-as-a-Service (MaaS) stealers.
This rule detects potential data exfiltration attempts or sensitive information disclosure through Microsoft Copilot Enterprise Search. It specifically looks for 'URLClick' actions from 'Microsoft Copilot Enterprise Search' where the raw event data contains a 'q=' parameter (indicating a query) and also contains keywords like 'extract', 'embed', or 'img'. This combination suggests that a user might be attempting to extract or embed content, potentially sensitive, via a Copilot search query, which could lead to data leakage.
Detects bulk file copies to removable drives for physical data exfiltration
Detects outbound web requests to Bing's 'Search by Image' endpoint where the embedded image URL points to a non-Microsoft domain. This mechanism is used in the SearchLeak attack (CVE-2026-42824) to bypass CSP and exfiltrate data from Microsoft 365 Copilot.
This rule detects network connections from endpoints to a list of known malicious domains. The detection is further refined by looking for specific URL paths such as '/wlc/', '/load/', or 'success/', which are commonly associated with command and control (C2) communication or data exfiltration activities by malware. This indicates potential compromise and active communication with adversary infrastructure.
Detects data exfiltration attempts using curl.exe to the public file transfer service sendit.sh. This rule specifically looks for the execution of 'curl.exe' with 'sendit.sh' in its command line arguments, indicating a potential attempt to upload or transfer data to this service.
This rule detects the explicit use of `curl` or `wget` to exfiltrate data, specifically AWS tokens, to a known Megalodon infrastructure IP address (216.126.225.129). This activity is indicative of the Megalodon CI/CD malware attempting to post extracted data.
Detects network events where a device communicates with the 'urlzzz.php' endpoint, specifically when this communication is correlated with file events involving 'Login Data' or 'Cookies'. This indicates potential exfiltration of browser credentials or session information by the UnregStealer malware.
This rule detects suspicious command-line activity that may indicate the leakage of API keys, specifically those resembling OpenAI API keys (starting with 'sk-', 'sk_live_', or 'sk_test_'). It looks for these patterns in conjunction with common data transfer utilities like curl, wget, powershell, or python, when executed by a process that is not a standard shell or explorer process. The rule triggers if two or more such events occur within an hour on the same computer.
Detects the execution of known ransomware-related executables ('servertool.exe', 'encrypt.exe') when their command line arguments indicate interaction with cloud storage services such as OneDrive, SharePoint, GoogleDrive, Dropbox, or iCloud. This behavior is indicative of ransomware attempting to encrypt or exfiltrate data from cloud environments.
Page 76 of 128