Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
472 detections
Filters
Last updated
All Time
Detection languages
342
74
19
16
7
Contributors
91
90
77
52
35
Categories
182
148
81
66
60
Platforms
406
68
63
58
26
Products / Services
10,680
9,761
6,694
4,418
3,744
472
MITRE Techniques
180
121
96
78
71
CVEs
6
5
5
5
2
IDS Classtypes
2
1
1
1
1
IDS Protocols
6
Detects a suspected EvilTokens attack sequence where a user clicks a phishing URL and subsequently initiates a browser session to the Microsoft device-code authorization page (devicelogin) within 15 minutes, with the phishing lure domain appearing in the browser context.
This rule detects high-risk administrative operations in Azure Active Directory (such as role assignment, password resets, and account status changes) that occur within 3 hours of a user's sign-in. This pattern often indicates an account compromise followed by immediate privilege escalation or persistence establishment.
The following analytic detects when a user in an Azure AD environment grants consent to an OAuth application. It leverages Azure AD audit logs to identify events where users approve application consents.
This activity is significant as it can expose organizational data to third-party applications, a common tactic used by malicious actors to gain unauthorized access.
If confirmed malicious, this could lead to unauthorized access to sensitive information and resources. Immediate investigation is required to validate the application's legitimacy, review permissions, and mitigate potential risks.
This activity is significant as it can expose organizational data to third-party applications, a common tactic used by malicious actors to gain unauthorized access.
If confirmed malicious, this could lead to unauthorized access to sensitive information and resources. Immediate investigation is required to validate the application's legitimacy, review permissions, and mitigate potential risks.
Detects anomalous database activity involving the creation of 'ZQ'-prefixed staging tables followed by a high-volume drop operation across multiple tables within a short timeframe. This behavior is indicative of data staging for potential exfiltration followed by an attempt to delete evidence or clear tracks.
Detects a multi-stage attack pattern against publicly exposed AI service endpoints. The detection correlates: (1) Enumeration of sensitive AI/MCP-related API endpoints (minimum 3 distinct hits) in a 10-minute window; (2) Subsequent POST requests to AI service endpoints within 15 minutes of the enumeration; and (3) A spike of 5 or more server errors (HTTP 5xx) within 10 minutes following the POST activity. This pattern suggests automated reconnaissance followed by attempted exploitation of exposed AI model interfaces or orchestration services. Covers T1595.002, T1190
Detects a multi-stage attack pattern against publicly exposed AI service endpoints. The detection correlates: (1) Enumeration of sensitive AI/MCP-related API endpoints (minimum 3 distinct hits) in a 10-minute window; (2) Subsequent POST requests to AI service endpoints within 15 minutes of the enumeration; and (3) A spike of 5 or more server errors (HTTP 5xx) within 10 minutes following the POST activity. This pattern suggests automated reconnaissance followed by attempted exploitation of exposed AI model interfaces or orchestration services. Covers T1595.002, T1190
Detects a multi-stage attack pattern against publicly exposed AI service endpoints. The detection correlates: (1) Enumeration of sensitive AI/MCP-related API endpoints (minimum 3 distinct hits) in a 10-minute window; (2) Subsequent POST requests to AI service endpoints within 15 minutes of the enumeration; and (3) A spike of 5 or more server errors (HTTP 5xx) within 10 minutes following the POST activity. This pattern suggests automated reconnaissance followed by attempted exploitation of exposed AI model interfaces or orchestration services. Covers T1595.002, T1190
Detects suspicious device registration or join events occurring within 60 minutes of a Device Code authentication flow. This pattern is consistent with the ARToken post-compromise technique, where an adversary uses an intercepted authentication code to register a new device to maintain persistent access to an Entra ID (Azure AD) environment.
Detects anomalous, high-frequency cloud management API requests across multiple sensitive paths (such as directory roles, applications, and service principals) from a single account. This pattern is indicative of automated enumeration or discovery of cloud infrastructure, permissions, and identity configurations, which is often a precursor to further exploitation.
Detects high-frequency, diverse Microsoft Graph API requests initiated by an identity, indicative of automated reconnaissance or enumeration of the cloud environment. The rule tracks requests across multiple categories such as organization settings, user/group identities, privilege information, application configurations, and resource access within a 30-minute window.
Detects high-volume, multi-endpoint enumeration activity against the Microsoft Graph API, involving a high frequency of requests across sensitive tenant resources (users, groups, applications, roles, mailboxes, etc.). This behavior is consistent with post-compromise reconnaissance by an adversary attempting to map the environment.
Detects network connection events involving a specific known malicious IP address (5.230.71.51) identified as participating in unauthorized Intune enrollment or Entra ID device registration processes.
Detects network connection events involving a specific known malicious IP address (5.230.71.51) identified as participating in unauthorized Intune enrollment or Entra ID device registration processes.
This detection identifies enabled user accounts that have not had a successful sign-in for the last 90 days, highlighting potential security risks from dormant accounts.
Detects when a user registers or adds a new device to Microsoft Entra ID (Azure AD) that has not been previously seen used by that user within the last 7 days. This behavior can indicate an adversary attempting to enroll an unauthorized device to bypass MFA or maintain persistent access.
This rule identifies instances where a user successfully authenticates using a phone app OTP MFA method that was registered between 1 and 30 days prior. The rule specifically flags logins associated with potentially automated or non-browser user agents, or logins flagged with risk, suggesting potential account takeover or persistence mechanisms involving MFA re-registration.
Detects anomalous, high-frequency access to multiple distinct Microsoft Graph API endpoints (e.g., users, groups, applications) by a single user within a short timeframe, which is indicative of tenant-wide reconnaissance following identity compromise.
This detection identifies enabled user accounts that have not had a successful sign-in for the last 90 days, highlighting potential security risks from dormant accounts.
Detects the creation or update of LiteLLM custom code guardrails that include suspicious Python snippets, specifically 'import os' and 'os.popen', which are indicative of attempts to execute arbitrary OS commands via guardrail logic (CVE-2026-59821).
Detects automated directory fuzzing attempts against Azure Storage blobs targeting sensitive files like .env, .pem, or configuration files. The rule identifies suspicious activity by aggregating requests from Anonymous or SAS-authenticated sources that exhibit high request volumes and access to multiple distinct sensitive paths, while excluding known legitimate bulk-transfer and synchronization tools.
Detects the use of the default, hardcoded 'sk-1234' secret key for HS256 JWT signing in LiteLLM proxy logs. This configuration is insecure and allows for the forgery of arbitrary user session tokens.
Page 7 of 23