Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
2,831 detections
Filters
Last updated
All Time
Detection languages
1,374
495
363
281
175
Contributors
848
453
393
252
188
Categories
20,440
11,814
5,871
5,145
4,864
2,831
Platforms
1,448
916
693
670
264
Products / Services
994
700
613
469
216
MITRE Techniques
1,239
911
484
481
388
CVEs
6
5
2
1
1
IDS Classtypes
15
10
6
5
3
IDS Protocols
39
Detects the deletion of a storage bucket immediately followed by the creation of a new bucket with the same name across different cloud providers (GCP, AWS S3, Azure). This rule specifically flags scenarios where the deleting actor or project differs from the creating actor or project, indicating a potential cross-account hijack or data exfiltration attempt.
Detects a potential S3 bucket hijacking attempt where an S3 bucket is deleted and shortly after, an AWS Firehose delivery stream is created or updated to target a bucket with the same name, but under a different AWS account. This could indicate an adversary attempting to exfiltrate data by redirecting data intended for the original bucket to their own controlled bucket.
This rule detects suspicious activities related to AWS S3 buckets that could indicate a bucket hijacking attempt or unauthorized replication to an external account. Specifically, it looks for two scenarios: 1) an S3 bucket being deleted and then recreated by a different AWS account within a one-hour window, suggesting a namespace hijacking, or 2) an S3 bucket being deleted while a PutBucketReplication event references an external AWS account, indicating an attempt to exfiltrate data or maintain access through replication.
Detects a sequence of events where an Amazon S3 bucket is deleted, followed within one hour by the creation of an Amazon Data Firehose delivery stream or an update to a destination using the same bucket name. This pattern indicates a potential bucket name hijacking attempt, where an attacker recreates the deleted bucket to intercept or redirect data streams processed by Firehose.
Detects a potential storage container hijacking attack in Azure, where a container is deleted in one subscription and subsequently recreated with the same name in a different subscription within the same tenant. This sequence can be used to redirect data streams or intercept access to containerized resources.
Detects potential GCP logging sink bucket hijacking by monitoring for the deletion of a Google Cloud Storage bucket followed closely by an update to a logging sink configuration that points to a GCS destination. This pattern is characteristic of attackers attempting to intercept cloud log streams by recreating a target destination bucket in an environment they control.
Detects bulk reconnaissance against multiple Microsoft Graph API endpoints (users, groups, roles, etc.) by scripting clients, correlated with a recent device registration event by the same user principal. This behavior is indicative of a Storm-2372 conditional access bypass strategy followed by rapid directory environment enumeration.
Detects anomalous Azure AD device registration events where device metadata is inconsistent, suggesting impersonation. The rule flags Linux hosts masquerading as Windows, non-Windows devices claiming AzureAD trust type, or the use of zeroed OS versions (0.0.0.0). High-frequency registrations (3+ devices per user in 30 minutes) correlate with known AADInternals-based attacks used by threat actor Storm-2372 to bypass Conditional Access policies.
Detects the use of Azure VM Run Command to execute PowerShell scripts that attempt to retrieve workload identity tokens from the Azure Instance Metadata Service (IMDS) endpoint. This activity could indicate an attacker attempting to steal credentials or escalate privileges within the Azure environment.
Detects cross-tenant service principal authentications that leverage blueprint-derived identities. This rule specifically looks for client_credentials grant types using client_assertion (JWT bearer) or client_secret targeting sensitive scopes like 'api://AzureADTokenExchange/.default' or 'AgentIdentity.CreateAsManager' from a foreign tenant service principal. This activity could indicate an attempt to establish persistence or gain unauthorized access across Azure AD tenants.
This rule detects when an application role assignment or delegated permission grant is successfully added to a service principal in Azure Active Directory, specifically looking for highly privileged permissions such as 'RoleManagement.ReadWrite.Directory', 'UserAuthenticationMethod.ReadWrite.All', 'UserAuthMethod-TAP.ReadWrite.All', or 'User.Read.All'. This could indicate an adversary attempting to gain elevated access or persistence within the Azure AD environment.
This rule detects the creation of Azure DNS zones by callers who have no prior history of writing to DNS zones within a specified lookback period. This behavior can indicate an attempt at DNS zone takeover, as observed in campaigns like 'Borrowed Trust' or 'link99', where adversaries might create orphaned DNS zones to redirect traffic or host malicious content.
This rule detects potential unauthorized access to Azure Storage Blobs following the retrieval of storage account access keys. It correlates 'ListKeys' API events with subsequent blob access activities from either unrecognized IP addresses or via Shared Access Signature (SAS) tokens, indicating potential credential misuse.
Detects successful retrieval of storage account access keys by a user identity rather than an established deployment or backup service.
Detects malicious npm package propagation indicative of the Miasma/Mini Shai-Hulud worm. The rule identifies suspicious patterns in AuditLogs, specifically burst npm publishes by known malicious actors or involving known malicious package names, as well as the use of weaponized GitHub Actions workflows or unauthorized snapshot branches.
This rule detects a spike in successful Azure Load Balancer operations from a single caller within an hour. A count of 3 or more successful operations from the same caller within a one-hour window will trigger this alert. This could indicate automated activity or unusual configuration changes related to load balancers.
Detects the creation of Kubernetes pods where the 'hostPID' or 'hostNetwork' settings are enabled. These configurations allow a pod to share the host's process namespace or network stack, respectively, which are common vectors for container escapes and host-level privilege escalation.
Detects the Miasma worm propagating through CI/CD pipelines by identifying unauthorized updates to GitHub Actions workflow files. Specifically monitors for modifications to workflows under the LeoPlatform or LeoInsights projects by the compromised user 'czirker', with signs of weaponization such as the addition of OIDC id-token write permissions and npm-publish capabilities.
Detects a single new session identifier accessing Exchange, SharePoint, and OneDrive within a short period, which may indicate broad post-authentication collection following an adversary-in-the-middle (AiTM) attack.
Detects attempts to modify or delete Conditional Access Policies in Microsoft Graph. This could indicate an adversary attempting to weaken security controls to facilitate unauthorized access or persistence.
Detects the creation of a Temporary Access Pass (TAP) by a service principal or agent identity, rather than a human administrator. This scenario indicates a potential compromise where a Temporary Access Agent with UserAuthMethod-TAP.ReadWrite.All permission is abused to generate a TAP for a Global Administrator.
Page 90 of 135