canvas
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides visualization capabilities using industry-standard tools and formats like Mermaid and draw.io.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests external code and data for diagramming, it implements defensive measures including structured verification steps, mandatory source citations for all diagram edges, and explicit marking of uncertain or inferred components to prevent data poisoning.
- [SAFE]: All external references and dependencies (e.g., Mermaid.js, D2, tldraw, Excalidraw) point to well-known, official documentation and reputable open-source projects.
Audit Metadata