Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
505 detections
Filters
Last updated
All Time
Detection languages
245
92
78
48
19
Contributors
141
113
53
46
41
Categories
20,456
11,839
5,876
5,155
4,865
Platforms
210
138
110
80
76
Products / Services
147
74
61
41
28
MITRE Techniques
243
128
69
68
55
CVEs
10
9
5
5
5
IDS Classtypes
4
IDS Protocols
4
Detects unauthorized attempts to disable, reconfigure, or bypass AI safety guardrails and moderation services, particularly when followed by high-risk tool invocations such as file deletion, network egress, or credential access. This activity indicates an attacker attempting to impair an AI agent's defensive mechanisms to facilitate malicious actions.
Detects unauthorized or suspicious modifications to AI model artifacts, including fine-tuned weights and LoRA adapters. This rule identifies tampering attempts via unsigned uploads, bypassing CI/CD production deployment approvals, or mismatches in hash/checksum compared to known-good versions.
Detects AI agents attempting to disable or bypass security guardrails, approval gates, or content filters. Additionally, it identifies high-risk or destructive actions taken by an AI agent that lack the required preceding human-approval audit record.
This rule monitors for unauthorized changes to GitHub Actions workflows that involve release or publishing processes, specifically looking for indicators of supply chain compromise like the introduction of 'id-token: write' or changes to publish scripts. It correlates these modifications with subsequent workflow executions and npm registry publish events attributed to the same actor to identify potential account takeover and malicious package distribution.
Detects high-privilege OAuth consent grants to applications shortly followed by suspicious mailbox modifications, such as creating new inbox or transport rules. This behavior is indicative of a post-compromise activity where an adversary uses an illicitly granted OAuth application to maintain persistence or exfiltrate data from an Exchange Online environment.
This rule detects suspicious activity associated with Redis server processes attempting to write to SSH 'authorized_keys' files or executing commands (such as BGREWRITEAOF or directory reconfiguration) commonly used by attackers to inject malicious SSH public keys into a Redis installation for persistent remote access.
Detects the creation or modification of inbox rules in Microsoft 365 that filter messages containing financial keywords (e.g., 'direct deposit', 'bank') and perform actions designed to hide or suppress these emails (e.g., 'MoveToFolder', 'DeleteMessage', 'MarkAsRead', 'StopProcessingRules'). This behavior is commonly used by adversaries to hide notification emails regarding financial transactions, which could alert the user to unauthorized activity.
Detects M365 administrative actions related to mailbox forwarding rules or OAuth application consent that occur in the absence of a Multi-Factor Authentication (MFA) claim, particularly when preceded by non-interactive session events (LogonType 0) using REST client protocols. This pattern is indicative of potential session hijacking or token replay attacks, where an adversary leverages a compromised session to establish persistence or gain ongoing access to the environment without re-authenticating.
Detects the creation of new Entra ID applications or service principals that use display names similar to well-known SaaS vendors but lack a verified publisher status, followed within 24 hours by a consent grant for high-privilege permissions. This pattern is characteristic of OAuth-impersonation campaigns aimed at gaining persistent access to cloud resources.
Detects OAuth application consent grant events where an application name matches common 'Data Loader' naming conventions and requests high-privilege scopes (read/write/full/api). This pattern is consistent with consent phishing campaigns used by threat actors to gain persistent access to SaaS environments like M365 or Salesforce.
Detects anomalous engineering account activity associated with supply-chain compromise, specifically identifying the generation of access tokens or retrieval of secrets followed by unauthorized changes to CI/CD pipeline configurations or secret stores in platforms like GitHub, JFrog Artifactory, or BrowserStack.
Detects anomalous engineering account activity associated with supply-chain compromise, specifically identifying the generation of access tokens or retrieval of secrets followed by unauthorized changes to CI/CD pipeline configurations or secret stores in platforms like GitHub, JFrog Artifactory, or BrowserStack.
Detects unauthorized modifications to AI Agent Model Context Protocol (MCP) tool manifests. This includes the injection of imperative instructions into tool descriptions, the addition of suspicious instruction-related fields to tool output schemas, or manifest updates initiated by unverified publishers.
This rule detects potentially malicious activity related to AI agent tool manipulation. It flags two primary scenarios: 1) The invocation of an AI tool shortly after an unmanaged or potentially unauthorized modification to its definition, and 2) Detection of prompt injection patterns within the output generated by an AI agent tool, which could indicate an attacker attempting to override the agent's instructions.
This rule detects potentially malicious activity related to AI agent tool manipulation. It flags two primary scenarios: 1) The invocation of an AI tool shortly after an unmanaged or potentially unauthorized modification to its definition, and 2) Detection of prompt injection patterns within the output generated by an AI agent tool, which could indicate an attacker attempting to override the agent's instructions.
Detects instances where a user grants high-privilege permissions (Mail.Read, Files.ReadWrite.All, offline_access) to an OAuth application within 24 hours of a risky sign-in event. This pattern is commonly observed in AiTM (Adversary-in-the-Middle) or consent-phishing attacks designed to establish persistent access to cloud resources.
Detects unauthorized modifications to the Entra ID authentication methods policy or suspicious application activity often associated with the staging of rogue External Authentication Methods (EAM). Adversaries may exploit these configuration changes to intercept or bypass multi-factor authentication (MFA) during user login processes.
Detects the suspicious reactivation of 'actions-cool/issues-helper' or 'actions-cool/maintain-one-comment' repositories within a known malicious window (September 14-19, 2026). These repositories were involved in the 'Mini Shai-Hulud' supply-chain attack. This rule identifies when previously disabled repositories are re-enabled or modified to become active, potentially re-exposing malicious release tags to downstream users.
Detects the suspicious reactivation of 'actions-cool/issues-helper' or 'actions-cool/maintain-one-comment' repositories within a known malicious window (September 14-19, 2026). These repositories were involved in the 'Mini Shai-Hulud' supply-chain attack. This rule identifies when previously disabled repositories are re-enabled or modified to become active, potentially re-exposing malicious release tags to downstream users.
Detects unauthorized modifications to AI agent configuration files, including system prompts, tool allow-lists, and MCP server registration files. Such changes are often performed by processes outside of approved configuration management or infrastructure-as-code pipelines, potentially allowing for persistent manipulation of AI model behavior and capabilities.
Detects potential post-authentication persistent actions (e.g., OAuth app grants, MFA changes, mailbox rule updates) performed shortly after a successful Google Workspace login, specifically flagging actions originating from IP addresses different from the login session to identify session hijacking or token replay (AiTM) activity. Includes suppression for known corporate IP ranges, trusted OAuth application consents, and recurring user behavior.
Page 7 of 25