Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
2,588 detections
Filters
Last updated
All Time
Detection languages
1,194
459
310
204
161
Contributors
487
455
323
171
123
Categories
1,884
937
415
400
360
Platforms
1,349
560
343
329
257
Products / Services
601
314
306
286
271
MITRE Techniques
18,115
15,635
12,577
8,388
6,220
2,588
CVEs
9
5
5
5
2
IDS Classtypes
142
96
49
11
6
IDS Protocols
177
49
44
17
4
Detects high-risk GitHub events indicating data or asset exposure, such as creating public repositories with specific naming conventions, changing private repositories to public, or creating new gists, which could indicate exfiltration of sensitive assets or information.
Detects a suspicious pattern where an AI coding agent or user creates a public GitHub repository with specific naming conventions (e.g., 'sweeper-demo', 'gitshot-images') and subsequently pushes content containing sensitive terms like 'screenshot' within a short timeframe. This behavior is indicative of potential data exfiltration of internal development screenshots to a public repository.
Detects a suspicious pattern where an AI coding agent or user creates a public GitHub repository with specific naming conventions (e.g., 'sweeper-demo', 'gitshot-images') and subsequently pushes content containing sensitive terms like 'screenshot' within a short timeframe. This behavior is indicative of potential data exfiltration of internal development screenshots to a public repository.
This rule detects anomalous spikes in network connections from a device to major LLM provider API endpoints (OpenAI, Azure OpenAI, Anthropic) by establishing a baseline for each device and identifying deviations exceeding a statistical threshold. Such behavior may indicate unauthorized data exfiltration or mass data processing using LLM services.
This detection identifies execution of the file synchronization utility "rclone".
It leverages Cisco Network Visibility Module logs, specifically flow data in order to capture process executions
initiating network connections.
While rclone is a legitimate command-line tool for syncing data to cloud storage providers, it has been widely abused by threat actors for data exfiltration.
This analytic inspects process name and arguments for rclone and flags usage of suspicious flags.
If matched, this could indicate malicious usage for stealthy data exfiltration or cloud abuse.
It leverages Cisco Network Visibility Module logs, specifically flow data in order to capture process executions
initiating network connections.
While rclone is a legitimate command-line tool for syncing data to cloud storage providers, it has been widely abused by threat actors for data exfiltration.
This analytic inspects process name and arguments for rclone and flags usage of suspicious flags.
If matched, this could indicate malicious usage for stealthy data exfiltration or cloud abuse.
This rule detects potential post-exploitation activity where a web server process spawns a command shell, executes system discovery commands, and immediately initiates an outbound network connection. This behavior is indicative of a web-based exploit leading to interactive command execution and subsequent C2 communication.
Detects unauthorized sharing, copying, or exporting of EBS, disk, or database snapshots to external cloud accounts, subscriptions, or projects that are not present in the defined trusted allow-list for AWS, Azure, or GCP environments.
Detects high-frequency outbound network connections to known large language model (LLM) API providers (OpenAI, xAI, Anthropic) originating from a single endpoint within a short timeframe. This activity is indicative of potential API key theft, where an attacker uses a compromised legitimate key to exhaust credit quotas or exfiltrate data via unauthorized API calls, bypassing the intended application.
Detects the use of cloud synchronization utilities such as rclone (including renamed binaries) or Mega command-line tools to interact with cloud storage services. The rule correlates this execution with preceding mass file enumeration or access activity on the same host, which is a common pattern in ransomware double-extortion scenarios.
Detects the installation of unauthorized AI-related browser extensions followed by outbound network communication from the browser process to known external AI service API endpoints within one hour. This behavior is indicative of potential data exfiltration via shadow AI tools, bypassing standard enterprise DLP controls.
Detects the installation of unauthorized AI-related browser extensions followed by outbound network communication from the browser process to known external AI service API endpoints within one hour. This behavior is indicative of potential data exfiltration via shadow AI tools, bypassing standard enterprise DLP controls.
Detects potential DNS tunneling or Command and Control (C2) activity by analyzing DNS query patterns. The rule identifies anomalous behavior based on high query volume, excessive unique subdomain fanout to a single parent domain, and unusually long average query lengths. This combination is often indicative of data exfiltration or communication with a C2 server over the DNS protocol.
Detects usage of the Rclone utility (or renamed binaries) to transfer data to various cloud storage backends (S3, Mega, pCloud, SFTP, WebDAV). The detection correlates this activity with the execution of common archival tools (7-Zip, WinRAR, TAR) within one hour prior to the data transfer, which is a common behavior of adversaries preparing data for exfiltration.
Detects outbound network traffic associated with the Azazel operation, including connections to identified C2 infrastructure, open directory/upload listeners, domain-specific staging servers, and exfiltration attempts to MEGA cloud storage endpoints.
Detects potential insider threats or departing employees by correlating abnormal bulk data downloads from cloud repositories (SharePoint/OneDrive/Box/etc.) with subsequent sensitive data access and exfiltration signals. It uses baseline behavioral profiling, requires a threshold of sensitive file interactions, and mandates high-confidence exfiltration indicators (removable media usage, suspicious external domains, or multiple combined destination events) to reduce noise. Archive/staging activity is used as an optional confidence booster.
Detects potential insider threats or departing employees by correlating abnormal bulk data downloads from cloud repositories (SharePoint/OneDrive/Box/etc.) with subsequent sensitive data access and exfiltration signals. It uses baseline behavioral profiling, requires a threshold of sensitive file interactions, and mandates high-confidence exfiltration indicators (removable media usage, suspicious external domains, or multiple combined destination events) to reduce noise. Archive/staging activity is used as an optional confidence booster.
Detects high-risk GitHub events indicating data or asset exposure, such as creating public repositories with specific naming conventions, changing private repositories to public, or creating new gists, which could indicate exfiltration of sensitive assets or information.
Detects the creation or exposure of public repositories, releases, or gists under a personal GitHub account that utilize naming conventions associated with the 'PixelLeak' workaround. This technique allows adversaries to make previously private screenshots, recordings, or internal assets publicly accessible by hosting them in public personal repositories or releases.
Detects anomalous, high-volume file download or export activity originating from a single identity across multiple downstream cloud applications or tenants within a one-hour window. This behavior is indicative of a compromised shared service provider, MSP, or clearinghouse account being leveraged to exfiltrate data from multiple client environments simultaneously. The detection excludes pre-approved service accounts and relies on a dynamic baseline comparison rather than a static threshold.
This rule detects unauthorized or anomalous outbound network connections from known AI agent runtime processes to LLM or API endpoints. It helps identify potential data exfiltration by monitoring for connections to destinations outside of an organization's approved allow-list for model providers, effectively catching scenarios where a compromised AI agent is prompted to relay sensitive data to an attacker-controlled endpoint.
Detects suspicious processes frequently polling Hugging Face repository endpoints (discussions, raw files, commits). This behavior is characteristic of adversaries using public code/dataset hosting services as a covert command-and-control (C2) channel to fetch instructions or exfiltrate data, bypassing traditional network filters by blending in with legitimate developer traffic.
Page 5 of 124