draw

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the soffice command-line tool to automate drawing and conversion tasks.
  • Evidence: Shell commands for headless conversion (soffice --headless --convert-to) and drawing (soffice --draw) in SKILL.md.
  • [PROMPT_INJECTION]: The skill processes untrusted external vector graphics and document files (SVG, PDF, ODG) which could contain malicious instructions designed to influence the agent.
  • Ingestion points: External files ingested via conversion commands and Python UNO scripts in SKILL.md.
  • Boundary markers: The instructions lack explicit boundary markers or warnings to disregard content embedded within processed files.
  • Capability inventory: The skill utilizes shell execution capabilities (soffice) and file system access to read and write document data (SKILL.md).
  • Sanitization: No sanitization or content validation steps are documented for the files being processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 09:39 AM
Security Audit — agent-trust-hub — draw