Daniel Jeremiah
@daniyyellUnited KingdomTrusted contributorCompletionist
8 followers4,600 downloads2,430 copies68 likes128,310 views
978 detections
Filters
Last updated
All Time
Detection languages
303
179
131
120
97
Categories
317
280
168
139
136
Platforms
669
130
98
61
39
Products / Services
112
54
50
48
38
MITRE Techniques
370
280
196
192
187
CVEs
35
7
4
4
4
IDS Classtypes
11
3
2
1
IDS Protocols
11
2
2
1
1
Detects an account granting offline access, mail, files, directory, or user impersonation permissions shortly after anomalous authentication. This rule is designed to identify potential persistence mechanisms established by adversaries leveraging OAuth applications, as seen in campaigns like Tycoon 2FA AITM.
Detects a browser creating a ZIP archive whose name impersonates a GPU, disk, temperature, codec, or PDF utility associated with high-performance systems. This activity is often associated with poisoned search results leading to cryptojacking campaigns.
Detects collection of service principals, app role assignments, delegated grants, and ownership data via Microsoft Graph API, which can identify privilege escalation paths. The rule specifically looks for a high volume of GET requests (35 or more within a 15-minute window) to URIs associated with enumerating service principals, app role assignments, oauth2 permission grants, and owned objects.
Detects creation and deletion of an Azure SQL firewall rule by the same caller within a short window, which can provide temporary database access while reducing later evidence. This behavior is often indicative of an adversary attempting to gain temporary access to a database, perform actions, and then remove their tracks by deleting the firewall rule.
Clipboard and screenshot staging pair
Cortex XDR
This rule detects the creation of 'Clipboard.txt' and 'Screenshot.bmp' files by the same process, indicating potential data staging activities. It leverages XQL to maintain actor and causality fields for investigation.
Detects a browser launching cmd.exe or PowerShell immediately after a user visits or downloads content presented as an AI coding assistant installer. This rule specifically looks for process creation events where a browser (Chrome, Edge, Firefox) is the parent process, and the child process is a command shell (cmd.exe, powershell.exe, pwsh.exe), with the command line containing keywords related to 'claude', 'install', 'setup', 'copy paste', 'mshta', or 'download'.
Detects a user or application issuing a high volume of Microsoft Graph user directory requests within ten minutes, consistent with tenant reconnaissance. This rule specifically looks for GET requests to the /v1.0/users endpoint of the Microsoft Graph API, indicating an attempt to enumerate user accounts.
Detects installation or update of a VM extension whose settings invoke scripts, remote support tooling, or command execution outside standard configuration management. This could indicate an adversary establishing persistence or remote access.
Detects JavaScript using the 'ghost_once_footer_' key with localStorage read and write operations, indicative of potential web-based attacks leveraging compromised Ghost CMS instances. This rule specifically looks for the string 'ghost_once_footer_' in conjunction with 'localStorage.getItem' and 'localStorage.setItem', suggesting a mechanism for persistent or one-time data storage in the browser's local storage.
Detects a compromised account revoking refresh tokens, signing out other sessions, or replacing authentication state to retain exclusive access. This behavior is indicative of an attacker attempting to maintain exclusive control over a compromised account by invalidating other legitimate sessions or tokens.
Page 4 of 98