icon-draw

Pass

Audited by Gen Agent Trust Hub on Jul 5, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill runs a local Python linter (scripts/check_svg.py) to validate SVG output. This script is contained within the skill and uses standard libraries, ensuring a secure validation process.
  • [DATA_EXFILTRATION]: There are no network-based operations or attempts to access sensitive credentials. The provided linter script explicitly blocks external URLs and raster images, mitigating data exfiltration risks via SVG assets.
  • [PROMPT_INJECTION]: The instructions in SKILL.md are focused on geometric constraints and design workflows. No patterns of safety bypass or instruction override were detected.
  • [REMOTE_CODE_EXECUTION]: The skill does not perform any remote code downloads or untrusted executions. All logic is self-contained within the skill's provided script and template files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 5, 2026, 11:16 PM
Security Audit — agent-trust-hub — icon-draw