analyzing-malware-persistence-with-autoruns

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: No malicious patterns or behaviors were detected during the analysis of the skill's code and documentation.
  • [COMMAND_EXECUTION]: The skill includes a Python script that utilizes subprocess.run to call the autorunsc64.exe binary. This command execution is required for the skill's primary function of scanning Windows systems for persistence mechanisms.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it processes untrusted external data from CSV reports.
  • Ingestion points: The scripts scripts/agent.py and the code block in SKILL.md ingest data from CSV files.
  • Boundary markers: Not present; the scripts read CSV columns directly as strings.
  • Capability inventory: Local file reading and command execution via subprocess.run in SKILL.md.
  • Sanitization: The data is parsed using standard libraries, but the content itself is not sanitized for potential injection strings targeting downstream LLM processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 03:56 PM