windows-shell-reliability

Pass

Audited by Gen Agent Trust Hub on Nov 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides documentation on how to reliably format and execute shell commands on Windows. It contains standard development practices for PowerShell and CMD environments.
  • [PROMPT_INJECTION]: The use of the term "CRITICAL" in section headers refers to technical reliability pitfalls (such as path quoting or encoding issues) rather than attempts to bypass agent safety guidelines or override system instructions.
  • [COMMAND_EXECUTION]: The provided code snippets demonstrate standard, safe usage of built-in Windows cmdlets (e.g., Remove-Item, Copy-Item, New-Item) and established developer tools like the .NET CLI and NPM. No malicious command strings or payload deliveries were detected.
  • [DATA_EXFILTRATION]: No network-enabled commands (such as curl or Invoke-WebRequest) are present. There are no references to sensitive file paths, credentials, or environment variables containing secrets.
  • [PRIVILEGE_ESCALATION]: The skill explicitly advises against using elevation as a default fix for permission errors, promoting safer troubleshooting practices.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 3, 2026, 06:07 AM