Windows PowerShell Script Execution via Redirected Input Stream

Flags PowerShell/pwsh executions where the command line includes redirected input ("- <").

FreeUnreviewedSigmahighv1
title: Windows PowerShell Script Execution via Redirected Input Stream
id: 12e50a99-c99c-439d-9a94-db9bebf8f19c
status: test
description: This rule identifies PowerShell or pwsh being launched with a command line containing a redirected input indicator ("- <"). Attackers can use redirected input to pass script content or commands through the process invocation, helping execution blend into less obvious command-line patterns. Detection relies on Windows process creation telemetry capturing the PowerShell executable path and the full command line.
references:
  - https://github.com/LOLBAS-Project/LOLBAS/blob/4db780e0f0b2e2bb8cb1fa13e09196da9b9f1834/yml/LOLUtilz/OSBinaries/Powershell.yml
  - https://twitter.com/Moriarty_Meng/status/984380793383370752
  - https://github.com/SigmaHQ/sigma/blob/master/rules/windows/process_creation/proc_creation_win_powershell_run_script_from_input_stream.yml
author: Moriarty Meng (idea), Anton Kutepov (rule), oscd.community, Huntrule Team
date: 2020-10-17
modified: 2021-11-27
tags:
  - attack.execution
  - attack.t1059
logsource:
  category: process_creation
  product: windows
detection:
  selection:
    Image|endswith:
      - \powershell.exe
      - \pwsh.exe
    CommandLine|re: \s-\s*<
  condition: selection
falsepositives:
  - Unknown
level: high
license: DRL-1.1
related:
  - id: c83bf4b5-cdf0-437c-90fa-43d734f7c476
    type: derived

What it detects

This rule identifies PowerShell or pwsh being launched with a command line containing a redirected input indicator ("- <"). Attackers can use redirected input to pass script content or commands through the process invocation, helping execution blend into less obvious command-line patterns. Detection relies on Windows process creation telemetry capturing the PowerShell executable path and the full command line.

Known false positives

  • Unknown

Detection content is published as a reviewed draft. Tune thresholds and exclusions against your own telemetry before enabling this rule for alerting.