draw-diagram

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Node.js scripts to orchestrate rendering, calls the d2 CLI for text-based diagram generation, and uses Playwright to launch a headless Chromium instance for processing Excalidraw scenes.
  • [EXTERNAL_DOWNLOADS]: The skill fetches icon assets from the official Iconify API (api.iconify.design) and downloads the Chromium browser binary when required by the Playwright rendering engine. These targets are well-known technology services and standard tools.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data which serves as a potential attack surface.
  • Ingestion points: Processes diagram source files (.excalidraw, .d2) and downloads SVG assets from api.iconify.design.
  • Boundary markers: Content is scoped to specific project directories (docs/src/, docs/assets/) and follows strict file naming conventions.
  • Capability inventory: Executes shell commands (node, d2, curl, npm), writes to the local filesystem, and launches a browser process.
  • Sanitization: The skill performs SVG filtering to reject scripts, event handlers, and external references. Additionally, the rendering environment is network-isolated, allowing traffic only to the local loopback address.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 02:43 PM
Security Audit — agent-trust-hub — draw-diagram