Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
4,543 detections
Filters
Last updated
All Time
Detection languages
2,478
532
344
301
286
Contributors
1,303
815
737
215
193
Categories
20,456
11,839
5,876
5,155
4,865
4,543
Platforms
3,841
1,049
473
39
29
Products / Services
1,835
1,138
211
183
171
MITRE Techniques
1,604
1,423
1,300
749
473
CVEs
19
17
16
11
11
IDS Classtypes
23
1
IDS Protocols
16
7
1
Baseline Comparison.
Author: miflower.
The purpose of this query is to perform a comparison between "known good" machines and suspected bad machines.
The original concept for this query was born due to reapplying the same 'whitelist' filters over and over.
It brings deltas between a baseline and another machine quickly to the analyst's view.
This query supports multiple suspected bad machines and multiple "known good" machines.
It also supports providing a timeframe for how far back in time to build a baseline as well as how far back in time to evaluate the suspected bad machines.
Each of the links provided by DeviceId/DeviceName will go to the most recent entry for whatever entity is listed.
Average results for the pre-defined settings below with a single good host and a single bad host on a 'huge' tenant (300k+ machines):.
Compute Time: ~10-20 seconds.
Result Set Size: ~500 rows.
The workflow is as follows:.
1. Establish Variables that are editable on a per-query basis.
2. Define functions for reuse.
3. Calculate DeviceIds for all machines in scope.
4. Derive deltas using the aforementioned functions.
5. Union together all results into a single view.
The following datasets are returned:.
1. Alerts on the suspected bad machines (ignores known good machines, because...they're alerts, additional data has the triggered file).
2. Connected Networks (from DeviceNetworkInfo table, additional data has full Connected Network details).
3. File Creations (disabled by default due to volume, enable at your own risk, additional data has initiating processes).
4. Image Loads (disabled by default due to volume, enable at your own risk, additional data has initiating processes).
5. Logon (derived from DeviceLogonEvents for the unique users logged on, additional data has logon types).
6. Network communication (grouped by 2nd level-domain, ie 'microsoft.com' in 'www.microsoft.com' and 'web.microsoft.com', additional data has the full list of URLs).
7. Process creation (additional da
Author: miflower.
The purpose of this query is to perform a comparison between "known good" machines and suspected bad machines.
The original concept for this query was born due to reapplying the same 'whitelist' filters over and over.
It brings deltas between a baseline and another machine quickly to the analyst's view.
This query supports multiple suspected bad machines and multiple "known good" machines.
It also supports providing a timeframe for how far back in time to build a baseline as well as how far back in time to evaluate the suspected bad machines.
Each of the links provided by DeviceId/DeviceName will go to the most recent entry for whatever entity is listed.
Average results for the pre-defined settings below with a single good host and a single bad host on a 'huge' tenant (300k+ machines):.
Compute Time: ~10-20 seconds.
Result Set Size: ~500 rows.
The workflow is as follows:.
1. Establish Variables that are editable on a per-query basis.
2. Define functions for reuse.
3. Calculate DeviceIds for all machines in scope.
4. Derive deltas using the aforementioned functions.
5. Union together all results into a single view.
The following datasets are returned:.
1. Alerts on the suspected bad machines (ignores known good machines, because...they're alerts, additional data has the triggered file).
2. Connected Networks (from DeviceNetworkInfo table, additional data has full Connected Network details).
3. File Creations (disabled by default due to volume, enable at your own risk, additional data has initiating processes).
4. Image Loads (disabled by default due to volume, enable at your own risk, additional data has initiating processes).
5. Logon (derived from DeviceLogonEvents for the unique users logged on, additional data has logon types).
6. Network communication (grouped by 2nd level-domain, ie 'microsoft.com' in 'www.microsoft.com' and 'web.microsoft.com', additional data has the full list of URLs).
7. Process creation (additional da
This query was originally published in the threat analytics report, CVE-2018-15982 exploit attacks.
CVE-2018-15982 is an exploit of Adobe Flash Player, that allows for remote execution of arbitrary code. It has since been patched.
Actors have been observed using this vulnerability in targeted attacks. Exploits for CVE-2018-15982 have also been included in several exploit kits.
In some initial attacks exploiting CVE-2018-15982, attackers sent targets spear-phishing emails. The emails would include an attached RAR archive, which contained a lure document, as well as a second archive disguised as a .jpg file. Opening the document would automatically run an embedded Flash ActiveX control. This, in turn, would call a script containing the exploit. The exploit's ability to run arbitrary code would be employed to unpack and run a payload from the second archive. The payload is a backdoor used both to achieve persistance and for command and control.
The following query detects possible instances of a payload being extracted by the exploit.
References:
https://nvd.nist.gov/vuln/detail/CVE-2018-15982
https://helpx.adobe.com/security/products/flash-player/apsb18-42.html
CVE-2018-15982 is an exploit of Adobe Flash Player, that allows for remote execution of arbitrary code. It has since been patched.
Actors have been observed using this vulnerability in targeted attacks. Exploits for CVE-2018-15982 have also been included in several exploit kits.
In some initial attacks exploiting CVE-2018-15982, attackers sent targets spear-phishing emails. The emails would include an attached RAR archive, which contained a lure document, as well as a second archive disguised as a .jpg file. Opening the document would automatically run an embedded Flash ActiveX control. This, in turn, would call a script containing the exploit. The exploit's ability to run arbitrary code would be employed to unpack and run a payload from the second archive. The payload is a backdoor used both to achieve persistance and for command and control.
The following query detects possible instances of a payload being extracted by the exploit.
References:
https://nvd.nist.gov/vuln/detail/CVE-2018-15982
https://helpx.adobe.com/security/products/flash-player/apsb18-42.html
Check all created files.
That does not have extension ps1, bat or cmd to avoid IT Pro scripts.
That are not copied to C:\ to detect all file share, external drive, data partition that are not allowed, etc.
This could help to detect malicious insider/user that has unencrypted data partition and that are using it to exfiltrate data even while removable devices & cloud storage is blocked.
That does not have extension ps1, bat or cmd to avoid IT Pro scripts.
That are not copied to C:\ to detect all file share, external drive, data partition that are not allowed, etc.
This could help to detect malicious insider/user that has unencrypted data partition and that are using it to exfiltrate data even while removable devices & cloud storage is blocked.
The query for malicious file creations via TOCTOU Vulnerability in Leading Endpoint Detection and Response (EDR) and Antivirus (AV) Solutions.
- Microsoft Defender (CVE-2022-37971)
- Defender for Endpoint (CVE-2022-37971)
- SentinelOne EDR
- TrendMicro Apex One (CVE-2022-45797)
- Avast Antivirus (CVE-2022-4173)
- AVG Antivirus (CVE-2022-4173)
- Microsoft Defender (CVE-2022-37971)
- Defender for Endpoint (CVE-2022-37971)
- SentinelOne EDR
- TrendMicro Apex One (CVE-2022-45797)
- Avast Antivirus (CVE-2022-4173)
- AVG Antivirus (CVE-2022-4173)
Microsoft has observed Bazacall using Cobalt Strike in order to move laterally to other machines on the network.
The following analytic detects potentially suspicious file copy operations targeting the System32 or SysWow64 directories as source, often indicative of malicious activity. It leverages data from Endpoint Detection and Response (EDR) agents, focusing on activity initiated by command-line tools like cmd.exe or PowerShell. This behavior is significant as it may indicate an attempt to evade defenses by copying an existing binary from the system directory and renaming it. If confirmed malicious, this activity could allow an attacker to execute code undetected and potentially leading to system compromise or further lateral movement within the network.
'Finds instances where a file uploaded to blob or file storage and it is seen on an endpoint by Microsoft Defender XDR.'
'Extracts MDATP Alert for a web shell being placed on the server and then enriches this event with information from W3CIISLog to idnetigy the Attacker that placed the shell'
Expanding on DeviceEvents output with Attack Surface Reduction (ASR) rule descriptions.
The ActionType values of the ASR events already explain what rule was matched and if it was audited or blocked.
However, it could still be useful to have a more human-friendly description in the results.
Also, this query is a good example for how you could define your own lookup tables and join with them.
The events in the DeviceEvents table contain a GUID for the various ASR rules rather than a full description of the rule.
This query will create a table which has the description for each ASR rule as per https://docs.microsoft.com/windows/security/threat-protection/windows-defender-exploit-guard/enable-attack-surface-reduction.
This table is then joined to the output of a query against the DeviceEvents table and shows a summary count of the events by the newly defined description.
This query shows the ability to use joins and custom dimension tables.
See https://docs.loganalytics.io/docs/Language-Reference/Tabular-operators/join-operator for more information on the join syntax.
For more questions on this query, feel free to ping @FlyingBlueMonki on twitter or mattegen@microsoft.com via email.
First lets start by creating a table of the rule descriptions to rule guids.
The ActionType values of the ASR events already explain what rule was matched and if it was audited or blocked.
However, it could still be useful to have a more human-friendly description in the results.
Also, this query is a good example for how you could define your own lookup tables and join with them.
The events in the DeviceEvents table contain a GUID for the various ASR rules rather than a full description of the rule.
This query will create a table which has the description for each ASR rule as per https://docs.microsoft.com/windows/security/threat-protection/windows-defender-exploit-guard/enable-attack-surface-reduction.
This table is then joined to the output of a query against the DeviceEvents table and shows a summary count of the events by the newly defined description.
This query shows the ability to use joins and custom dimension tables.
See https://docs.loganalytics.io/docs/Language-Reference/Tabular-operators/join-operator for more information on the join syntax.
For more questions on this query, feel free to ping @FlyingBlueMonki on twitter or mattegen@microsoft.com via email.
First lets start by creating a table of the rule descriptions to rule guids.
'This hunting query will identify where a file is uploaded to Azure File or Blob storage
and is then accessed once before being deleted. This activity may be indicative of
exfiltration activity.'
and is then accessed once before being deleted. This activity may be indicative of
exfiltration activity.'
This query was originally published in the threat analytics report, Ryuk ransomware. There is also a related blog.
Ryuk is human-operated ransomware. Much like DoppelPaymer ransomware, Ryuk is spread manually, often on networks that are already infected with Trickbot.
Ryuk operators use PsExec to manually spread the ransomware to other devices.
The following query detects remote file creation events that might indicate an active attack.
The See also section below lists links to other queries associated with Ryuk ransomware.
References:
https://www.microsoft.com/security/blog/2020/03/05/human-operated-ransomware-attacks-a-preventable-disaster/
https://www.microsoft.com/wdsi/threats/malware-encyclopedia-description?Name=Ransom:Win32/Ryuk&threatId=-2147232689
https://www.microsoft.com/security/blog/2020/03/05/human-operated-ransomware-attacks-a-preventable-disaster/
https://docs.microsoft.com/sysinternals/downloads/psexec
Ryuk is human-operated ransomware. Much like DoppelPaymer ransomware, Ryuk is spread manually, often on networks that are already infected with Trickbot.
Ryuk operators use PsExec to manually spread the ransomware to other devices.
The following query detects remote file creation events that might indicate an active attack.
The See also section below lists links to other queries associated with Ryuk ransomware.
References:
https://www.microsoft.com/security/blog/2020/03/05/human-operated-ransomware-attacks-a-preventable-disaster/
https://www.microsoft.com/wdsi/threats/malware-encyclopedia-description?Name=Ransom:Win32/Ryuk&threatId=-2147232689
https://www.microsoft.com/security/blog/2020/03/05/human-operated-ransomware-attacks-a-preventable-disaster/
https://docs.microsoft.com/sysinternals/downloads/psexec
This query identifies files that are copied to a device over SMB, then executed within a
specified threshold. Default is 5 seconds, but is configurable by tweaking the value for
ToleranceInSeconds.
specified threshold. Default is 5 seconds, but is configurable by tweaking the value for
ToleranceInSeconds.
This query looks for signs of impacket psexec module usage. May hit other psexec-like techniques too.
Author: Jouni Mikkola
More info: https://threathunt.blog/impacket-psexec/
Author: Jouni Mikkola
More info: https://threathunt.blog/impacket-psexec/
This query looks for signs of credential dumping based on process activity instead of targeting process names.
Author: Jouni Mikkola
More info: https://threathunt.blog/lsass-credential-dumping/
Author: Jouni Mikkola
More info: https://threathunt.blog/lsass-credential-dumping/
Look for links opened from outlook.exe, followed by a browser download and then a SmartScreen app warning that was ignored by the user.
Read more about these events and this hunting approach in this post: https://techcommunity.microsoft.com/t5/forums/editpage/board-id/WDATPActor/message-id/34.
Data availability: SmartScreen events are available only on Windows 10 version 1703 and onwards.
Tags: #EmailLink, #BrowserDownload, #SmartScreen.
Read more about these events and this hunting approach in this post: https://techcommunity.microsoft.com/t5/forums/editpage/board-id/WDATPActor/message-id/34.
Data availability: SmartScreen events are available only on Windows 10 version 1703 and onwards.
Tags: #EmailLink, #BrowserDownload, #SmartScreen.
Use this query to locate persistence in Startup with HTA files.
First query digs in print spooler drivers folder for any file creations,
MANY OF THE FILES THAT SHOULD COME UP HERE MAY BE LEGIT. Unsigned files
or ones that don't have any relations to printers that you are using are
suspicious.
Second query that can be used for finding client machines that
could be operating print servers or file servers is also included here.
As additional mitigation for the exploit you might want to block the
incoming traffic to the SMB or EPMAP Ports (445) if you need to keep the
spooler service running to print from clients.
MANY OF THE FILES THAT SHOULD COME UP HERE MAY BE LEGIT. Unsigned files
or ones that don't have any relations to printers that you are using are
suspicious.
Second query that can be used for finding client machines that
could be operating print servers or file servers is also included here.
As additional mitigation for the exploit you might want to block the
incoming traffic to the SMB or EPMAP Ports (445) if you need to keep the
spooler service running to print from clients.
First query digs in print spooler drivers folder for any file creations,
MANY OF THE FILES THAT SHOULD COME UP HERE MAY BE LEGIT. Unsigned files
or ones that don't have any relations to printers that you are using are
suspicious.
Second query that can be used for finding client machines that
could be operating print servers or file servers is also included here.
As additional mitigation for the exploit you might want to block the
incoming traffic to the SMB or EPMAP Ports (445) if you need to keep the
spooler service running to print from clients.
MANY OF THE FILES THAT SHOULD COME UP HERE MAY BE LEGIT. Unsigned files
or ones that don't have any relations to printers that you are using are
suspicious.
Second query that can be used for finding client machines that
could be operating print servers or file servers is also included here.
As additional mitigation for the exploit you might want to block the
incoming traffic to the SMB or EPMAP Ports (445) if you need to keep the
spooler service running to print from clients.
'Dev-0530 actors are known to encrypt the contents of the victims device as well as renaming the file extensions. This query looks for the creation of files with .h0lyenc extension or presence of ransom note.'
This query looks for lnk file executions from other locations than C: -drive, which can relate to mounted ISO-files.
Reference - https://threathunt.blog/detecting-a-payload-delivered-with-iso-files-using-mde/
Reference - https://threathunt.blog/detecting-a-payload-delivered-with-iso-files-using-mde/
This query was originally published in the threat analytics report, EvilQuest signals the rise of Mac ransomware.
As of the time of this writing (October 2020), ransomware designed to target macOS is relatively rare. EvilQuest is one of the few examples of this kind of malware on the platform.
The query below can detect the creation of a ransom note according to the typical methods of EvilQuest operators. The command the query searches for is associated with, but not definitely indicative of, EvilQuest infections.
Other queries related to EvilQuest ransomware can be found under the See also section below.
As of the time of this writing (October 2020), ransomware designed to target macOS is relatively rare. EvilQuest is one of the few examples of this kind of malware on the platform.
The query below can detect the creation of a ransom note according to the typical methods of EvilQuest operators. The command the query searches for is associated with, but not definitely indicative of, EvilQuest infections.
Other queries related to EvilQuest ransomware can be found under the See also section below.
Page 204 of 217