Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
1,588 detections
Filters
Last updated
All Time
Detection languages
830
199
132
116
107
Contributors
608
289
249
88
63
Categories
20,440
11,814
5,871
5,145
4,864
1,588
Platforms
1,255
461
295
76
68
Products / Services
530
382
139
113
85
MITRE Techniques
629
565
504
437
297
CVEs
5
4
3
3
3
IDS Classtypes
5
2
2
1
IDS Protocols
5
3
2
Detects the 'comet.exe' process (Perplexity AI agent) performing file read, write, rename, or creation operations outside of standard browser paths (e.g., AppData, Temp, Downloads). This behavior potentially indicates the process is being leveraged for unauthorized data access or staging files on the local filesystem.
Detects unauthorized access to common browser credential and cookie files by processes other than standard web browsers. This activity is a common indicator of credential harvesting malware or unauthorized data collection.
Detects potential path traversal or unauthorized access attempts by Semantic Kernel Agents (or related processes like Python/DotNet) to sensitive host files. The rule monitors for file read, creation, or modification activities targeting sensitive credentials, configuration files (e.g., .ssh, .aws, .kube), and system files, which could indicate a sandbox escape or malicious data collection.
Detects unauthorized processes (non-browser) reading sensitive browser session and credential files (e.g., Cookies, Login Data) across multiple user profiles, followed by immediate outbound network connections. This behavior is highly characteristic of commodity information stealers (e.g., LummaC2, Stealc, Vidar) attempting to exfiltrate browser-stored credentials and session data.
Detects the invocation of file-related functions (DownloadFileAsync, upload_file) within the Semantic Kernel framework that lack associated validation function calls. This pattern is indicative of potential Remote Code Execution (RCE) or unauthorized file system access scenarios where an LLM agent is coerced into performing dangerous file operations without path sanitization.
Detects instances where a Semantic Kernel AI agent, running via Python or .NET, accesses sensitive local files like SSH keys, cloud credentials, or environment files immediately before triggering a potential file upload operation.
Detects instances where non-browser processes (e.g., cmd, powershell, or malicious binaries) attempt to access sensitive browser files such as 'Login Data' or 'Local State'. This is a common technique used by credential stealers to exfiltrate stored browser passwords, cookies, and configuration data.
Detects high-volume file access or download operations in SharePoint or OneDrive performed by common scripting libraries and client user agents, which may indicate automated data exfiltration.
Detects unauthorized processes attempting to read or manipulate local storage, session storage, or database files associated with messaging applications like Telegram and WhatsApp, or email clients like Outlook and Thunderbird. This behavior is indicative of credential or data theft by malicious software or unauthorized scripts.
Detects non-browser processes accessing sensitive browser-stored credential and cookie databases concurrently with access to Discord local storage, which is a common behavior of information-stealing malware (e.g., x47.c) to aggregate credentials for exfiltration.
Detects anomalous, mass-access behavior targeting plaintext credential files, certificate stores, private keys, or password-manager databases located on general-purpose network shares. Such activity is often indicative of an adversary conducting discovery or collection of credentials for lateral movement and privilege escalation.
Detects suspicious processes accessing cryptocurrency wallet data files or browser extension settings, characteristic of infostealer malware such as AMOS or GhostCode. The rule excludes common browser processes and legitimate wallet applications to identify unauthorized attempts to harvest wallet credentials, private keys, or seed phrases.
This rule detects a suspicious sequence of events indicative of a Redis Lua sandbox escape. It monitors for a network connection to Redis (port 6379) followed shortly by the redis-server process reading the system's /etc/passwd file, which is a known indicator of an attempted arbitrary file read via the EVAL command.
Detects a process accessing browser credential files (Login Data) followed by an outbound network connection to a suspicious exfiltration endpoint commonly associated with credential-stealing malware.
This rule detects processes other than known web browsers (Chrome, Edge, Firefox, Brave, Opera) accessing sensitive browser-related files (Login Data, Cookies, Web Data) and initiating an external network connection within 15 minutes. This behavior is highly characteristic of credential and session cookie theft, often utilized by info-stealing malware for subsequent session hijacking.
Detects SVG files containing embedded ECMAScript that executes a client-side redirect to a known AiTM phishing or C2 domain, specifically targeting the Mirage2FA delivery mechanism.
Detects instances where a non-browser process accesses multiple distinct browser credential or session artifacts (such as Login Data, Cookies, or Local State) across different browser profiles in a short period. This behavior is indicative of credential harvesting activities, often associated with C2 frameworks or post-exploitation tools attempting to steal session cookies, 2FA backup codes, or saved passwords.
Detects rapid, bulk cloning or downloading of GitHub repositories containing sensitive data (e.g., consensus algorithms, PII, internal blocklists). This behavior is characteristic of unauthorized data exfiltration, specifically tracking activity involving suspicious OAuth token prefixes (gho_) and rapid high-volume access to sensitive organizational assets within a short timeframe.
Detects instances where browser processes (Chrome or Edge) are spawned by smartscreen.exe and subsequently access sensitive browser storage files such as Login Data, Cookies, or Local State. This pattern is indicative of a process injection or masquerading chain used to extract credentials or session data.
Detects SVG image files that contain embedded HTML or script payloads, often used as an initial delivery vector to smuggle malicious code (such as downloaders or web redirects) into a target environment. This technique has been associated with the delivery of various remote access trojans (RATs).
Detects anomalous activity in a CI/CD environment where a runner process accesses sensitive credential files (such as PyPI tokens or GitHub secrets) followed closely by a network connection to public package registry domains, potentially indicating a supply-chain compromise where stolen credentials are used to publish malicious packages.
Page 14 of 76