Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
15,636 detections
Filters
Last updated
All Time
Detection languages
6,581
4,571
999
788
723
Contributors
3,615
2,087
2,043
1,738
1,271
Categories
9,459
6,773
3,297
2,511
1,598
Platforms
13,080
2,270
1,179
493
256
Products / Services
3,812
3,376
1,976
1,254
704
MITRE Techniques
18,116
15,636
12,583
8,393
6,223
CVEs
22
21
20
19
19
IDS Classtypes
175
58
53
25
11
IDS Protocols
256
53
17
8
4
Detects Windows Terminal or related console hosts spawning PowerShell or PowerShell Core with command-line arguments indicative of staging, downloading, or executing remote scripts (such as encoded commands, Invoke-Expression, or web requests), commonly associated with ClickFix or social engineering techniques.
This rule performs a sweep across device file and network events for indicators of compromise (IOCs) associated with the UAT-11587/Antino campaign. It detects malicious file hashes, specific C2 domain connections, and known lure URLs (HTA/WSF) identified by Cisco Talos.
Detects the deployment or creation of containers (via Docker, Podman, Nerdctl, Crictl, or Kubectl) with elevated risk flags such as privileged mode, host namespace sharing, root mounts, docker socket access, or initiated from web application parent processes.
Detects potential web shell activity where a web or application server process (such as IIS, Apache, Nginx, or Tomcat) writes a script file (e.g., .aspx, .jsp, .php) into a web-accessible directory, followed shortly by spawning a command-line shell or utility process.
Detects potential web shell activity where a web or application server process (such as IIS, Apache, Nginx, or Tomcat) writes a script file (e.g., .aspx, .jsp, .php) into a web-accessible directory, followed shortly by spawning a command-line shell or utility process.
Detects endpoint execution of specific known brute-force scripts and tools (such as ssh_brute.py, brute3.py, auto_login.py, etc.) using Python or shell interpreters, correlated with a high outbound network connection fan-out (50 or more distinct target IPs) from the same process.
Hunts for sqlservr.exe spawning command shells or system utilities, a pattern consistent with xp_cmdshell abuse. It also flags later steps from a public intrusion write-up: certutil decoding, SAM, SYSTEM or SECURITY hive saves and local account or group changes.
Results are grouped per device and day and ranked by stage count. Rows can also appear without a SQL parent when two later stages occur on the same device and day.
Triage: Some jobs use xp_cmdshell legitimately. Review the command line and the account. Hunting query, not an alert.
Limitations: Activity inside the SQL process with no child process is not visible. Renamed tools, obfuscation and privilege escalation tools are not covered. Needs Defender for Endpoint on the SQL servers.
Results are grouped per device and day and ranked by stage count. Rows can also appear without a SQL parent when two later stages occur on the same device and day.
Triage: Some jobs use xp_cmdshell legitimately. Review the command line and the account. Hunting query, not an alert.
Limitations: Activity inside the SQL process with no child process is not visible. Renamed tools, obfuscation and privilege escalation tools are not covered. Needs Defender for Endpoint on the SQL servers.
Detects commands executed on a Barracuda load balancer. Network appliances sit
outside most endpoint monitoring, so their own audit trail is frequently the only
record that anyone logged in and did something.
Onboarding: BARRACUDA-LOADBALANCER is ingested via Syslog (CEF) (Bloo extractor 859).
Integration guide: https://bloo.io/docs/hypercloud/syslog-connector
BQL (Bloo Query Language):
stream=auditd where sourcename='BARRACUDA-LOADBALANCER' and action='SESSION_ENDED'
| groupby user, process, commandline
outside most endpoint monitoring, so their own audit trail is frequently the only
record that anyone logged in and did something.
Onboarding: BARRACUDA-LOADBALANCER is ingested via Syslog (CEF) (Bloo extractor 859).
Integration guide: https://bloo.io/docs/hypercloud/syslog-connector
BQL (Bloo Query Language):
stream=auditd where sourcename='BARRACUDA-LOADBALANCER' and action='SESSION_ENDED'
| groupby user, process, commandline
Detects a Slack workflow being published. Workflows run automation inside the
workspace — they can post, collect responses, and reach external endpoints via
webhooks — which makes a malicious one a credential-harvesting form or an
exfiltration channel that looks like an internal tool.
Onboarding: SLACK is ingested via Slack native connector (Bloo extractor 833).
Integration guide: https://bloo.io/docs/integrations/slack
BQL (Bloo Query Language):
stream=slack-workflow where sourcename='SLACK' and action='PUBLISHED'
| groupby user, workflow
workspace — they can post, collect responses, and reach external endpoints via
webhooks — which makes a malicious one a credential-harvesting form or an
exfiltration channel that looks like an internal tool.
Onboarding: SLACK is ingested via Slack native connector (Bloo extractor 833).
Integration guide: https://bloo.io/docs/integrations/slack
BQL (Bloo Query Language):
stream=slack-workflow where sourcename='SLACK' and action='PUBLISHED'
| groupby user, workflow
Detects a privileged command executed through Centrify on Windows. Brokered
elevation records the command under the identity that requested it rather than the
shared administrative account it ran as, which restores attribution that local
Windows auditing frequently loses.
Onboarding: CENTRIFY-AUDITTRAIL-WINDOWS is ingested via Syslog (KV) (Bloo extractor 821).
Integration guide: https://bloo.io/docs/hypercloud/syslog-connector
BQL (Bloo Query Language):
stream=iam where sourcename='CENTRIFY-AUDITTRAIL-WINDOWS'
and action='USER_COMMAND_EXECUTED'
| groupby user, targetuser, role, domain
elevation records the command under the identity that requested it rather than the
shared administrative account it ran as, which restores attribution that local
Windows auditing frequently loses.
Onboarding: CENTRIFY-AUDITTRAIL-WINDOWS is ingested via Syslog (KV) (Bloo extractor 821).
Integration guide: https://bloo.io/docs/hypercloud/syslog-connector
BQL (Bloo Query Language):
stream=iam where sourcename='CENTRIFY-AUDITTRAIL-WINDOWS'
and action='USER_COMMAND_EXECUTED'
| groupby user, targetuser, role, domain
Detects a renamed PSF-signed pythonw.exe (with an original file name of pythonw.exe but a different on-disk name) launched from a user-writable path under Users. The command line consists only of its own path, which is consistent with ._pth based path configuration abuse for loading attacker code stealthily.
Detects the creation or modification of web script files (.aspx, .jsp) within common web application directories by unusual processes (such as command shells or download utilities) or by non-standard user accounts, indicating potential web shell deployment.
This rule performs a sweep across device file and network events for indicators of compromise (IOCs) associated with the UAT-11587/Antino campaign. It detects malicious file hashes, specific C2 domain connections, and known lure URLs (HTA/WSF) identified by Cisco Talos.
Detects when the IIS worker process (w3wp.exe) spawns command-line utilities or script interpreters such as cmd.exe, powershell.exe, mshta.exe, wscript.exe, or cscript.exe. This behavior often indicates web application exploitation or webshell activity.
Detects Java processes spawning PowerShell or Command Prompt to modify Windows Defender settings by adding exclusions for sensitive paths or extensions (such as AppData, Temp, Users Public, or ProgramData).
Detects endpoint execution of specific known brute-force scripts and tools (such as ssh_brute.py, brute3.py, auto_login.py, etc.) using Python or shell interpreters, correlated with a high outbound network connection fan-out (50 or more distinct target IPs) from the same process.
Detects endpoint execution of specific known brute-force scripts and tools (such as ssh_brute.py, brute3.py, auto_login.py, etc.) using Python or shell interpreters, correlated with a high outbound network connection fan-out (50 or more distinct target IPs) from the same process.
Hunts for NetScaler Pitboss packet engine messages where the text after NSPPE carries shell syntax such as separators, command substitution or ${IFS}. Security vendors observed this pattern during exploitation of the NetScaler pre authentication command injection flaw patched in September 2026.
The query reads Syslog and CommonSecurityLog and decodes URL encoding. It checks only the text after NSPPE, so plain heartbeat lines without shell syntax do not match. The matching logic was checked on sample strings only, not on real appliance logs.
Triage: Check the source, the appliance version and whether it was patched at that time. A match on an unpatched appliance points to an exploitation attempt, not proof of success. Preserve logs and a snapshot before upgrading. Hunting query, not an alert.
Limitations: Needs NetScaler syslog forwarded to Sentinel. Empty results do not prove an appliance is clean. Attackers can remove traces and other techniques are not covered. Field names depend on the syslog format and connector. Bare separators and short words can match legitimate messages, so tune the regex after reviewing real results.
The query reads Syslog and CommonSecurityLog and decodes URL encoding. It checks only the text after NSPPE, so plain heartbeat lines without shell syntax do not match. The matching logic was checked on sample strings only, not on real appliance logs.
Triage: Check the source, the appliance version and whether it was patched at that time. A match on an unpatched appliance points to an exploitation attempt, not proof of success. Preserve logs and a snapshot before upgrading. Hunting query, not an alert.
Limitations: Needs NetScaler syslog forwarded to Sentinel. Empty results do not prove an appliance is clean. Attackers can remove traces and other techniques are not covered. Field names depend on the syslog format and connector. Bare separators and short words can match legitimate messages, so tune the regex after reviewing real results.
Detects the creation of cook.pth or cook.py in Python site-packages or dist-packages directories by package managers or interpreters, followed by a Python interpreter execution on the same device within one hour. This pattern matches the malicious supply chain behavior observed in compromised LiteLLM versions (e.g., 1.107.0) utilizing .pth auto-import chains for persistence and execution.
Detects creation of a scheduled task named "Windows Defender Advanced
Threat Protection Service" configured to run at system startup with SYSTEM privileges
and launch an executable named defender.exe. HoneyMyte (Mustang Panda) used this exact
schtasks command to establish boot persistence for the CoolClient backdoor sideloader,
masquerading the task and binary as a legitimate Windows Defender component.
Threat Protection Service" configured to run at system startup with SYSTEM privileges
and launch an executable named defender.exe. HoneyMyte (Mustang Panda) used this exact
schtasks command to establish boot persistence for the CoolClient backdoor sideloader,
masquerading the task and binary as a legitimate Windows Defender component.
Detects the default scheduled task activity of the hacktool Defendnot, which is used to disable Windows Defender Security.
It abuses Microsoft's undocumented WSC API to register itself as a fake antivirus product, allowing it to disable Windows Defender.
During installation, Defendnot creates a scheduled task named 'defendnot' with the argument '--from-autorun' for persistence.
It abuses Microsoft's undocumented WSC API to register itself as a fake antivirus product, allowing it to disable Windows Defender.
During installation, Defendnot creates a scheduled task named 'defendnot' with the argument '--from-autorun' for persistence.
Page 8 of 745