Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
1,551 detections
Filters
Last updated
All Time
Detection languages
774
293
112
103
93
Contributors
493
267
209
73
60
Categories
1,319
1,227
492
174
155
Platforms
1,529
116
56
10
7
Products / Services
545
446
229
102
91
MITRE Techniques
18,123
15,639
12,581
8,394
6,215
1,551
CVEs
8
6
4
4
4
IDS Classtypes
1
1
IDS Protocols
1
1
KQL detects suspicious child processes launched by PHP or web-server workers following possible exploitation of CVE-2026-109618 or CVE-2026-109619.
Detects Node.js processes spawning shells, script hosts or download tools, which is how post exploitation of a Node.js server would look after CVE-2026-111497 (Brotli decoder out of bounds write, fixed in 22.23.4, 24.22.1 and 26.12.1). Devices that Defender Vulnerability Management lists as vulnerable are flagged and sorted first, and rows are still returned when the CVE is not listed. It excludes npm, npx, yarn and pnpm processes and GitHub Actions runner action processes, which start shells for build scripts. It does not detect the memory corruption itself. The query ran without errors in a live tenant.
Detects Windows web/PHP worker processes (php.exe, php-cgi.exe,
apache.exe, httpd.exe, lsphp.exe, nginx.exe) spawning unexpected command shells
or download utilities. Exploitation of CVE-2026-109618 / CVE-2026-109619 in core
PHP grants unauthenticated RCE; the child process is the observable artifact since
the memory corruption leaves no trace in the PHP logs. mod_php deployments are
the highest-risk configuration.
apache.exe, httpd.exe, lsphp.exe, nginx.exe) spawning unexpected command shells
or download utilities. Exploitation of CVE-2026-109618 / CVE-2026-109619 in core
PHP grants unauthenticated RCE; the child process is the observable artifact since
the memory corruption leaves no trace in the PHP logs. mod_php deployments are
the highest-risk configuration.
Detects web server worker processes or PHP interpreters spawning suspicious child processes associated with reverse shells, network sockets, or reconnaissance commands.
Detects instances where Microsoft Teams (ms-teams.exe, Teams.exe) or its associated WebView2 process spawns a command or script interpreter (such as PowerShell, Cmd, Mshta, Wscript, Cscript) with suspicious download-and-execute arguments, which is commonly associated with ClickFix social engineering attacks or similar initial access tactics.
Detects cmd.exe launching curl.exe with TLS verification disabled, redirect following, an output file, and forced hostname-to-IP resolution, or the PowerShell Invoke-WebRequest with a certificate-check bypass switch, often associated with malicious trojan droppers.
Detects potential staging and execution of files from the Microsoft Vault directory within AppData, often associated with adversary tooling or credential theft.
Detects AnyDesk being silently installed with the --install and --silent flags or having an unattended-access password configured via the --set-password and _unattended_access parameters, typically piped via stdin from a batch dropper using cmd.exe.
Detects Microsoft Word (winword.exe) spawning the Windows command shell (cmd.exe) to execute batch scripts (.bat or .cmd files). This behavior is commonly associated with malicious document macros attempting to drop and execute secondary payloads, excluding common system paths to reduce false positives.
Detects command-line execution of script interpreters (PowerShell, CMD, MSHTA) spawned via Windows Explorer containing download primitives and URLs, followed by immediate external network connections.
Detects WinRAR or 7-Zip extracting an archive with a hard-coded password supplied via command line from a cmd.exe batch parent, using switch patterns associated with the RareWolf AnyDesk dropper.
Detects instances where a browser process (Chrome, Edge, Firefox, Brave) is observed on a device shortly before an explorer.exe-initiated PowerShell, PWSH, or CMD process executes a command containing download strings or script execution patterns (such as DownloadString, Invoke-WebRequest, IEX, or encoded commands) referencing URLs.
MSI-backed loader execution chain
Cortex XDR
This query uses the process ID of each child to match the parent process's PID, then reconstructs the chain back to explorer.exe. It also checks for the encoded PowerShell, silent MSI installation, and embedded MSI execution behaviours described in the source article.
Important caveats
- Causality scope: This finds matching process events within a shared causality ID; it does not establish that each process directly launched the next.
- Explorer root: The root process may not share the same causality ID in every telemetry scenario. If this join removes otherwise valid results, investigate the root separately.
- Strict matching: Requiring every stage will miss partial chains and variants with different command-line arguments. Consider separate alerts for the encoded PowerShell and silent MSI behaviours.
- Validation: Confirm the field names, join syntax, and causality-ID population against your tenant before deploying this as a production detection.
Important caveats
- Causality scope: This finds matching process events within a shared causality ID; it does not establish that each process directly launched the next.
- Explorer root: The root process may not share the same causality ID in every telemetry scenario. If this join removes otherwise valid results, investigate the root separately.
- Strict matching: Requiring every stage will miss partial chains and variants with different command-line arguments. Consider separate alerts for the encoded PowerShell and silent MSI behaviours.
- Validation: Confirm the field names, join syntax, and causality-ID population against your tenant before deploying this as a production detection.
Detects when AI coding agent processes (such as Claude, Cursor, Copilot, Windsurf, Codeium, or Aider) spawn command shells or download utilities with command lines indicative of downloading and executing external payloads.
Detects execution and network activity associated with the cache_base operator toolkit, including JuicyPotatoNG (kitanai.exe), RegPwn (nandemo.exe), ufw.exe reverse shell, and upd-ms.bat launching these tools with connections to *.strangled.net on port 7070.
Detects observable hiding actions targeting ScreenConnect (ConnectWise Control) clients, such as modifying registry keys to hide uninstall entries, using attrib.exe to set hidden attributes, or launching management tools with hidden window styles.
Detects cmd.exe executing a ping-based delay to loopback followed by self-deletion of the batch script (%~f0), as well as bulk deletion of common archive and executable files, typically associated with malware droppers like RareWolf.
Detects when AppLocker "Audit only" enforcement mode reports that an Application, DLL, Script, MSI, or Packaged-App would have been blocked if AppLocker "Enforce rules" enforcement mode was enabled.
Detects when common office applications (Word, Excel, PowerPoint, Outlook, OneNote, etc.) or archive utilities/PDF readers spawn command-line interpreters or scripting engines (cmd.exe, powershell.exe, wscript.exe, etc.), which is a common indicator of phishing document execution or malicious archive extraction.
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 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.
Page 1 of 74