Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
66 detections
Filters
Last updated
All Time
Detection languages
18
16
11
8
7
Contributors
26
15
3
3
3
Categories
22
12
10
10
8
Platforms
20
19
16
13
10
Products / Services
28
11
9
7
6
MITRE Techniques
18,114
15,633
12,576
8,388
6,215
IDS Classtypes
8
2
1
IDS Protocols
7
3
1
Detects an inbound Microsoft Teams call or chat using a known STAC4749 helpdesk-impersonation persona display name, a caller UPN on a known STAC4749 .top IT-themed domain, or a source IP previously observed originating STAC4749 vishing calls.
Detects outbound network connections to identified infrastructure associated with 'Pink' (CL-CRI-1147) passkey-enrollment vishing campaigns. These operators use phone-based social engineering to direct victims to phishing domains that relay credentials and MFA tokens in real-time to register attacker-controlled passkeys. The rule monitors for specific DNS query patterns containing passkey-related keywords and known malicious IP addresses used by the threat group.
Detects a near-simultaneous correlation between a user registering a new FIDO2 authentication method and a login session from an unmanaged or non-compliant device. This behavioral pattern is often associated with adversary-in-the-middle or vishing attacks where an attacker tricks a user into registering an attacker-controlled security key to maintain persistent MFA access.
Detects the registration of a new FIDO2 security key or passkey in Entra ID that is associated with a sign-in from an unmanaged or non-compliant device shortly before or after the registration. This behavior often suggests a vishing or social engineering scenario where an attacker guides a user through an authentication method registration process while impersonating IT support.
Detects renamed AutoHotKey binaries launched from user-writable paths by a browser or Teams parent process, spawning cmd or powershell as children. UNC6692 ("Snow Flurries") impersonates IT helpdesk over Teams, lures victims to a fake "Mailbox Repair Utility" page, and delivers a renamed AHK binary + same-named script from an attacker S3 bucket to install the SNOWBELT extension.
Detects the execution of processes associated with voice manipulation, voice changing, or 'deepfake' software. These tools may be used by adversaries to facilitate social engineering, impersonation, or to bypass voice-based authentication mechanisms.
Detects the execution of .crx Edge extension files where the parent process is Microsoft Teams, often indicating social engineering attacks delivering malicious extensions.
Detects Payouts King ransomware operator TTPs involving Microsoft Teams social engineering followed by Windows Quick Assist (msra.exe) abuse. The rule monitors for three primary indicators: Teams spawning Quick Assist, Quick Assist being launched with remote session-initiation flags, and Quick Assist spawning suspicious child processes (cmd, powershell, etc.) or network utilities (curl, certutil, bitsadmin) characteristic of post-exploitation payload delivery.
Detects network connections to privnote.com used by UNC3753 for delivering installation commands during social engineering attacks. The threat actors use this self-destructing text service to transmit cURL commands and RMM installer links without leaving permanent browser footprints.
Detects when collaboration applications (like Microsoft Teams or Zoom) or web browsers launch remote support utilities. This activity may indicate social engineering tactics during a screen sharing session, where an attacker convinces a user to install or run remote access software.
Detects renamed AutoHotKey binaries launched from user-writable paths by a browser or Teams parent process, spawning cmd or powershell as children. This activity is associated with UNC6692 ("Snow Flurries") impersonating IT helpdesk over Teams, luring victims to a fake "Mailbox Repair Utility" page, and delivering a renamed AHK binary + same-named script from an attacker S3 bucket to install the SNOWBELT extension.
Detects Microsoft Teams spawning child processes (cmd, powershell, wscript, mshta, rundll32, regsvr32, certutil) after receiving external communications, a technique used by MuddyWater for initial access and execution.
This rule detects the download of common remote access tools (AnyDesk, TeamViewer, RustDesk, Zoho, Atera, UltraViewer) over HTTP. The detection is triggered when an HTTP response flows to the client, is established, contains a URI, and the URI matches one of the specified tool names using a case-insensitive PCRE. Additionally, the response body size must be greater than 100,000 bytes, indicating a potential executable or installer download. This activity is classified as trojan activity and is associated with the MITRE ATT&CK technique T1566.004 (Phishing: Voice Phishing) and tactic Initial Access, suggesting that the download might occur after a voice phishing call.
This rule detects HTTP traffic that appears to be a fake IT helpdesk page attempting to host a remote tool. It looks for the keywords "helpdesk", "download", and "remote" (case-insensitive) within the file data of HTTP responses. This combination of keywords suggests a social engineering attempt to trick users into downloading and executing remote access software under the guise of IT support.
This rule detects web pages that are likely part of a bank vishing callback scam. It looks for HTTP responses (flow:to_client) containing the keywords 'call', 'toll-free', and 'account' (case-insensitive) in the file data. This combination of keywords suggests a page designed to trick users into calling a fraudulent number to discuss their 'account'.
Detects cURL command used by UNC3753 to download SuperOps RMM agent during social engineering attacks. The threat actor instructs victims to execute this specific command pattern to install persistent remote access tools for data theft operations.
Detects potential SMS phishing (smishing) vectors involving a specific shortened URL used to deliver mobile surveillance and exfiltration malware, specifically the fake RedAlert application.
Identifies successful Entra ID sign-ins using the Device Code Flow from a geographic location not seen for that user in the last 30 days. The device code flow involves two separate interactions (the attacker's device and the victim's browser), so a mismatch between the sign-in location and the user's established locations is a strong indicator of a session hijacking attempt.
Identifies successful Entra ID sign-ins using the Device Code Flow from a geographic location not seen for that user in the last 30 days. The device code flow involves two separate interactions (the attacker's device and the victim's browser), so a mismatch between the sign-in location and the user's established locations is a strong indicator of a session hijacking attempt.
Detects when an administrator or help desk agent resets a user's password or deactivates/resets their MFA factors. This activity, while often legitimate, is a key tactic used by social engineering groups like Muddled Libra (aka Scattered Spider) who are known to 'vishing' (voice-phishing) IT help desks to take over accounts.
Detects suspicious activities in Salesforce environments consistent with the TTPs of the ShinyHunters threat group (also tracked as UNC6040/UNC6240). The attack involves social engineering (vishing) to trick users into authorizing malicious OAuth applications that impersonate legitimate tools like 'Salesforce Data Loader'. This rule looks for key indicators such as the installation of unreviewed packages, creation of impersonating connected apps, elevation of permissions to 'full' API access, and large data exports.
Page 3 of 4