EXPLORE DETECTIONS
MCP Prompt Injection
This detection identifies potential prompt injection attempts within MCP (Model Context Protocol) communications by monitoring for known malicious phrases and patterns commonly used to manipulate AI assistants. Prompt injection is a critical vulnerability where adversaries embed hidden instructions in content processed by AI tools, attempting to override system prompts, bypass security controls, or hijack the AI's behavior. The search monitors JSON-RPC traffic for phrases such as "IGNORE PREVIOUS INSTRUCTIONS," "SYSTEM PROMPT OVERRIDE," and "ignore all security" which indicate attempts to subvert the AI's intended behavior and potentially execute unauthorized actions through the MCP toolchain.
MCP Sensitive System File Search
This detection identifies MCP filesystem tool usage attempting to search for files containing sensitive patterns such as passwords, credentials, API keys, secrets, and configuration files. Adversaries and malicious insiders may abuse legitimate MCP filesystem capabilities to conduct reconnaissance and discover sensitive data stores for exfiltration or credential harvesting.
Microsoft Defender ATP Alerts
The following analytic is to leverage alerts from Microsoft Defender ATP Alerts. This query aggregates and summarizes all alerts from Microsoft Defender ATP Alerts, providing details such as the source, file name, severity, process command line, ip address, registry key, signature, description, unique id, and timestamps. This detection is not intended to detect new activity from raw data, but leverages Microsoft provided alerts to be correlated with other data as part of risk based alerting. The data contained in the alert is mapped not only to the risk object, but also the threat object. This detection filters out evidence that has a verdict of clean from Microsoft. It dynamically maps the MITRE technique at search time to auto populate the annotation field with the value provided in the alert. It also uses a dynamic mapping to set the risk score in Enterprise Security based on the severity of the alert.
Microsoft Defender Incident Alerts
The following analytic is to leverage alerts from Microsoft Defender O365 Incidents. This query aggregates and summarizes all alerts from Microsoft Defender O365 Incidents, providing details such as the destination, file name, severity, process command line, ip address, registry key, signature, description, unique id, and timestamps. This detection is not intended to detect new activity from raw data, but leverages Microsoft provided alerts to be correlated with other data as part of risk based alerting. The data contained in the alert is mapped not only to the risk object, but also the threat object. This detection filters out evidence that has a verdict of clean from Microsoft. It dynamically maps the MITRE technique at search time to auto populate the annotation field with the value provided in the alert. It also uses a static mapping to set the risk score based on the severity of the alert.
Microsoft Intune Bulk Wipe
The following analytic detects a high volume of "wipe ManagedDevice" events from the Intune admin portal (5+ per hour by default). It leverages Intune audit logs to identify when this action is triggered. This activity is significant beacuse the "wipe ManagedDevice" action factory resets devices connected to your Microsoft Intune tenant. If confirmed malicious, an attacker can abuse this action to perform a large-scale data wiping attack against your managed endpoints.
Microsoft Intune Device Health Scripts
Microsoft Intune device remediation scripts are a tool administrators can use to remotely manage devices, this functionality can also be abused for SYSTEM level code execution and lateral movement to intune managed devices. This detection identifies when a new device health script has been added, updated or deleted.
Microsoft Intune DeviceManagementConfigurationPolicies
Microsoft Intune device management configuration policies are a tool administrators can use to remotely manage policies and settings on intune managed devices. This functionality can also be abused to disable defences & evade detection. This detection identifies when a new device management configuration policy has been created.
Microsoft Intune Manual Device Management
Microsoft Intune device management configuration policies, scripts & apps are a all tools administrators can use to remotely manage intune managed devices. Instead of waiting for the devices to poll for changes to polciies, the policies can be manually pushed to expidite delivery. This may be useful in a pinch, it may also be a sign of an impatient attacker trying to speed up the delivery of their payload. This detection identifies when a device management configuration policy sync events, on-demand remediation scripts are triggered or when devices are remotely restarted.
Microsoft Intune Mobile Apps
Microsoft Intune supports deploying packaged applications to support software deployment, this functionality can also be abused for deploying malicious payloads to intune managed devices. This detection identifies when a new packaged application has been added, updated or deleted.
Microsoft SharePoint Server Elevation of Privilege
The following analytic detects potential exploitation attempts against Microsoft SharePoint Server vulnerability CVE-2023-29357. It leverages the Web datamodel to monitor for specific API calls and HTTP methods indicative of privilege escalation attempts. This activity is significant as it may indicate an attacker is trying to gain unauthorized privileged access to the SharePoint environment. If confirmed malicious, the impact could include unauthorized access to sensitive data, potential data theft, and further compromise of the SharePoint server, leading to a broader security breach.
Mimikatz PassTheTicket CommandLine Parameters
The following analytic detects the use of Mimikatz command line parameters associated with pass-the-ticket attacks. It leverages data from Endpoint Detection and Response (EDR) agents, focusing on specific command-line patterns related to Kerberos ticket manipulation. This activity is significant because pass-the-ticket attacks allow adversaries to move laterally within an environment using stolen Kerberos tickets, bypassing normal access controls. If confirmed malicious, this could enable attackers to escalate privileges, access sensitive information, and maintain persistence within the network.
Mmc LOLBAS Execution Process Spawn
The following analytic identifies `mmc.exe` spawning a LOLBAS execution process. It leverages data from Endpoint Detection and Response (EDR) agents, focusing on process creation events where `mmc.exe` is the parent process. This activity is significant because adversaries can abuse the DCOM protocol and MMC20 COM object to execute malicious code, using Windows native binaries documented by the LOLBAS project. If confirmed malicious, this behavior could indicate lateral movement, allowing attackers to execute code remotely, potentially leading to further compromise and persistence within the environment.
Modification Of Wallpaper
The following analytic detects the modification of registry keys related to the desktop wallpaper settings. It leverages Sysmon EventCode 13 to identify changes to the "Control Panel\\Desktop\\Wallpaper" and "Control Panel\\Desktop\\WallpaperStyle" registry keys, especially when the modifying process is not explorer.exe or involves suspicious file paths like temp or public directories. This activity is significant as it can indicate ransomware behavior, such as the REVIL ransomware, which changes the wallpaper to display a ransom note. If confirmed malicious, this could signify a compromised machine and the presence of ransomware, leading to potential data encryption and extortion.
Modify ACL permission To Files Or Folder
The following analytic detects the modification of ACL permissions to files or folders, making them accessible to everyone or to system account. It leverages data from Endpoint Detection and Response (EDR) agents, focusing on processes like "cacls.exe," "icacls.exe," and "xcacls.exe" with specific command-line arguments. This activity is significant as it may indicate an adversary attempting to evade ACLs or access protected files. If confirmed malicious, this could allow unauthorized access to sensitive data, potentially leading to data breaches or further system compromise.
Monitor Email For Brand Abuse
The following analytic identifies emails claiming to be sent from a domain similar to one you are monitoring for potential abuse. It leverages email header data, specifically the sender's address, and cross-references it with a lookup table of known domain permutations generated by the "ESCU - DNSTwist Domain Names" search. This activity is significant as it can indicate phishing attempts or brand impersonation, which are common tactics used in social engineering attacks. If confirmed malicious, this could lead to unauthorized access, data theft, or reputational damage.
Monitor Registry Keys for Print Monitors
The following analytic detects modifications to the registry key `HKLM\SYSTEM\CurrentControlSet\Control\Print\Monitors`. It leverages data from the Endpoint.Registry data model, focusing on events where the registry path is modified. This activity is significant because attackers can exploit this registry key to load arbitrary .dll files, which will execute with elevated SYSTEM permissions and persist after a reboot. If confirmed malicious, this could allow attackers to maintain persistence, execute code with high privileges, and potentially compromise the entire system.
Monitor Web Traffic For Brand Abuse
The following analytic identifies web requests to domains that closely resemble your monitored brand's domain, indicating potential brand abuse. It leverages data from web traffic sources, such as web proxies or network traffic analysis tools, and cross-references these with known domain permutations generated by the "ESCU - DNSTwist Domain Names" search. This activity is significant as it can indicate phishing attempts or other malicious activities targeting your brand. If confirmed malicious, attackers could deceive users, steal credentials, or distribute malware, leading to significant reputational and financial damage.
MOVEit Certificate Store Access Failure
This detection identifies potential exploitation attempts of the CVE-2024-5806 vulnerability in Progress MOVEit Transfer. It looks for log entries indicating failures to access the certificate store, which can occur when an attacker attempts to exploit the authentication bypass vulnerability. This behavior is a key indicator of attempts to impersonate valid users without proper credentials. While certificate store access failures can occur during normal operations, an unusual increase in such events, especially from unexpected sources, may indicate malicious activity.
MOVEit Empty Key Fingerprint Authentication Attempt
This detection identifies attempts to authenticate with an empty public key fingerprint in Progress MOVEit Transfer, which is a key indicator of potential exploitation of the CVE-2024-5806 vulnerability. Such attempts are characteristic of the authentication bypass technique used in this vulnerability, where attackers try to impersonate valid users without providing proper credentials. While occasional empty key fingerprint authentication attempts might occur due to misconfigurations, a sudden increase or attempts from unexpected sources could signify malicious activity. This analytic helps security teams identify and investigate potential exploitation attempts of the MOVEit Transfer authentication bypass vulnerability.
MS Exchange Mailbox Replication service writing Active Server Pages
The following analytic identifies the creation of suspicious .aspx files in specific directories associated with Exchange exploitation by the HAFNIUM group and the ProxyShell vulnerability. It detects this activity by monitoring the MSExchangeMailboxReplication.exe process, which typically does not write .aspx files. This behavior is significant as it may indicate an active exploitation attempt on Exchange servers. If confirmed malicious, attackers could gain unauthorized access, execute arbitrary code, or maintain persistence within the environment. Immediate investigation and remediation are crucial to prevent further compromise.
MS Scripting Process Loading Ldap Module
The following analytic detects the execution of MS scripting processes (wscript.exe or cscript.exe) loading LDAP-related modules (Wldap32.dll, adsldp.dll, adsldpc.dll). It leverages Sysmon EventCode 7 to identify these specific DLL loads. This activity is significant as it may indicate an attempt to query LDAP for host information, a behavior observed in FIN7 implants. If confirmed malicious, this could allow attackers to gather detailed Active Directory information, potentially leading to further exploitation or data exfiltration.
MS Scripting Process Loading WMI Module
The following analytic detects the loading of WMI modules by Microsoft scripting processes like wscript.exe or cscript.exe. It leverages Sysmon EventCode 7 to identify instances where these scripting engines load specific WMI-related DLLs. This activity is significant because it can indicate the presence of malware, such as the FIN7 implant, which uses JavaScript to execute WMI queries for gathering host information to send to a C2 server. If confirmed malicious, this behavior could allow attackers to collect sensitive system information and maintain persistence within the environment.
MSBuild Suspicious Spawned By Script Process
The following analytic detects the suspicious spawning of MSBuild.exe by Windows Script Host processes (cscript.exe or wscript.exe). This behavior is often associated with malware or adversaries executing malicious MSBuild processes via scripts on compromised hosts. The detection leverages Endpoint Detection and Response (EDR) telemetry, focusing on process creation events where MSBuild is a child of script hosts. This activity is significant as it may indicate an attempt to execute malicious code. If confirmed malicious, it could lead to unauthorized code execution, potentially compromising the host and allowing further malicious activities.
Mshta spawning Rundll32 OR Regsvr32 Process
The following analytic detects a suspicious mshta.exe process spawning rundll32 or regsvr32 child processes. It leverages data from Endpoint Detection and Response (EDR) agents, focusing on process GUID, process name, and parent process fields. This activity is significant as it is a known technique used by malware like Trickbot to load malicious DLLs and execute payloads. If confirmed malicious, this behavior could allow attackers to execute arbitrary code, escalate privileges, or download additional malware, posing a severe threat to the environment.