fsecure-internet-security-malware-distribution

Fail

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: HIGHCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [DATA_EXFILTRATION]: The skill provides instructions to harvest sensitive user application data.
  • Evidence: The command Copy-Item $historyPath "$env:USERPROFILE\Desktop\chrome_history_backup.db" instructs the agent to copy the Google Chrome browser history database (which contains URLs, search terms, and timestamps) to a predictable location.
  • [COMMAND_EXECUTION]: The skill includes commands for destructive file operations on the host system.
  • Evidence: The command Remove-Item -Force is used with broad patterns (*fsecure*, *f-secure*) in the user's Downloads directory, creating a risk of accidental deletion of legitimate files.
  • [COMMAND_EXECUTION]: The skill provides commands for deep system reconnaissance and persistence checking.
  • Evidence: Commands include reg query for registry-based persistence, Get-ScheduledTask for auditing scheduled tasks, and netstat -ano for inspecting active network connections.
  • [COMMAND_EXECUTION]: The skill provides instructions to execute system security utilities and process management tools.
  • Evidence: Commands include executing MpCmdRun.exe for Windows Defender scans and using tasklist with findstr to monitor specific running processes.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jul 6, 2026, 10:37 AM
Security Audit — agent-trust-hub — fsecure-internet-security-malware-distribution