Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
3,867 detections
Filters
Last updated
All Time
Detection languages
2,526
525
299
218
131
Contributors
999
706
554
263
188
Categories
1,449
1,199
1,091
793
576
Platforms
39,960
7,270
6,537
3,867
3,651
Products / Services
1,756
1,449
1,121
438
406
MITRE Techniques
1,889
1,392
913
814
727
CVEs
12
5
4
3
2
IDS Classtypes
10
7
6
6
5
IDS Protocols
37
2
1
This rule normalizes common Azure AD sign-in fields such as user principal names, source IP addresses, application display names, and authentication requirements for consistent monitoring and analysis.
Detects the reuse of a session or refresh token within a short timeframe (<= 900 seconds) where the source IP location, device identifier, or User-Agent string has materially changed. This behavior is indicative of Adversary-in-the-Middle (AiTM) attacks, where a session cookie has been stolen and replayed from an attacker's infrastructure to hijack a legitimate user's session.
Detects activity from known Storm-3168 (JADEPUFFER) campaign IP addresses across Azure control-plane, sign-in, and App Service logs. The rule monitors for malicious indicators in Azure Activity logs, Azure Sign-in logs, and Service Principal sign-ins, as well as specific sensitive URI paths (e.g., shells, admin login paths) within App Service HTTP logs.
Detects rapid, large-scale deletion of Azure storage accounts by a single identity within a short timeframe. This behavior is indicative of destructive activity, such as a malicious actor attempting to disrupt service availability or impact organizational data, rather than typical administrative teardown.
This rule detects reconnaissance activity targeting Azure App Service instances, specifically looking for attempts to access web-shell-like files, WordPress administration paths, PHP-CGI endpoints, or LangFlow code validation paths. It identifies potential Storm-3168 actor infrastructure through known IOCs or through patterns of high-volume, path-diverse HTTP requests that deviate from typical noise, helping to distinguish targeted probing from common internet background scanning.
This rule detects reconnaissance activity targeting Azure App Service instances, specifically looking for attempts to access web-shell-like files, WordPress administration paths, PHP-CGI endpoints, or LangFlow code validation paths. It identifies potential Storm-3168 actor infrastructure through known IOCs or through patterns of high-volume, path-diverse HTTP requests that deviate from typical noise, helping to distinguish targeted probing from common internet background scanning.
Detects anomalous service principal activity characterized by multiple, rapidly issued, and overlapping OAuth tokens within a short time window. This pattern is associated with automated/agentic token orchestration (e.g., JADEPUFFER/Storm-3168 campaign) rather than standard manual operator activity, specifically when leveraging the python-requests user agent.
Detects sign-in events within Microsoft Entra ID (formerly Azure AD) that utilize the 'deviceCode' authentication protocol. This OAuth 2.0 flow is designed for devices with limited input (e.g., smart TVs, IoT devices, CLI tools) but is frequently abused by attackers in 'device code phishing' campaigns to obtain access tokens by tricking users into entering a code on a malicious site.
Detects anomalous activity where a successful device-code authentication is immediately followed by rapid multi-device registration or Intune enrollment. This pattern is consistent with GhostCode-style attacks involving Primary Refresh Token (PRT) theft and rogue device registration to bypass conditional access policies.
Detects instances where a privileged account (such as an administrator or helpdesk user) removes MFA security information or resets a password for a target account, followed shortly (within 4 hours) by a sign-in attempt from that same target account exhibiting risk characteristics such as impossible travel or unfamiliar features. This pattern is often associated with social engineering attacks, such as those attributed to Scattered Spider, where an attacker coerces support staff to disable MFA protections for an account.
Detects potential Adversary-in-the-Middle (AiTM) activity by identifying a successful interactive MFA-protected sign-in followed within 15 minutes by a non-interactive (refresh token) sign-in for the same user from a different IP address, ASN, or User-Agent. This pattern is indicative of session hijacking where an attacker reuses a stolen session token.
Detects a sequence of events where a user with a high-risk or compromised sign-in status registers a new device in Entra ID shortly after the event, followed by subsequent sign-ins from that device that bypass multi-factor authentication (MFA) via Conditional Access policies.
Detects MFA relay phishing kits (such as Bluekit) where the MFA challenge is satisfied by an attacker-controlled proxy rather than the legitimate user device. The rule identifies sign-in attempts where the primary authentication request originates from one network location, while the MFA completion step originates from a different network location, and the MFA claim indicates it was satisfied via a proxied token.
Detects the registration or modification of multi-factor authentication (MFA) security information in Entra ID or Okta that occurs within 15 minutes of a sign-in event flagged as risky. This behavioral pattern is a common indicator of persistence being established by an attacker after performing an adversary-in-the-middle (AiTM) attack, where they leverage a stolen session to register their own MFA factors to maintain long-term access.
This rule detects the unauthorized download of complete ML model checkpoint or weight files from a model registry or object storage. It monitors for common model file extensions in download events that are not associated with a legitimate deployment justification or a pre-approved change management window, helping to identify potential intellectual property theft or model exfiltration attempts.
Detects high-privilege user accounts (e.g., executives, administrators) undergoing a credential or MFA reset, followed by an atypical sign-in event within a two-hour window. This pattern is indicative of potential account takeover or identity-compromise scenarios, often associated with help-desk vishing attacks where an attacker induces a user or administrator to reset credentials or MFA methods.
Detects a two-stage activity: first, the potential unauthorized reading or access of LLM provider API keys from files, environment variables, or command-line arguments; and second, the subsequent use of those same credentials to authenticate to LLM provider APIs (OpenAI, Anthropic, Azure, Bedrock) from a different, unrecognized source IP address, indicating potential credential theft and session hijacking.
Detects Azure AD / Entra ID sign-in events that utilize the OAuth 2.0 device code authentication flow. This protocol, designed for input-constrained devices, is frequently exploited by adversary-in-the-middle phishing kits (e.g., EvilTokens) to trick users into authorizing a malicious session, effectively bypassing traditional MFA by obtaining an active session token.
Detects instances where a user authenticates via the OAuth 2.0 device code flow followed shortly by a new device registration. The rule flags registrations where the device name shape is uncommon compared to historical data or where there is a mismatch between the IP address used for the authentication and the registration, indicating potential device code phishing or account takeover.
Detects anomalous sign-in patterns indicative of real-time MFA relay attacks, where a user principal authenticates successfully from multiple distinct IP addresses and geographical locations within a short time window (15 minutes), suggesting the active interception of credentials and MFA challenges by a Phishing-as-a-Service platform.
Detects high-volume, rapid enumeration of core Microsoft Graph directory endpoints (users, groups, applications, and servicePrincipals) by a single identity or session. This behavior is indicative of reconnaissance tools such as AADInternals, ROADtools, or TokenTactic being used to map out a cloud environment's directory structure.
Page 23 of 185