excalidraw-diagram

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the Excalidraw library from the well-known esm.sh CDN and provides instructions for downloading the Chromium browser binary via Playwright.
  • [COMMAND_EXECUTION]: Employs a Python script (render_excalidraw.py) that manages headless browser processes to automate diagram rendering.
  • [DYNAMIC_EXECUTION]: Evaluates Javascript within a sandboxed headless browser environment to process JSON data and export SVG/PNG assets.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is designed to ingest and visualize external data (e.g., API specifications, protocol documentation).
  • Ingestion points: SKILL.md (Research mandates requiring the agent to ingest external documentation and data formats).
  • Boundary markers: Absent (no specific delimiters to separate untrusted external data from the generator's instructions).
  • Capability inventory: render_excalidraw.py (Capability to write files and execute code within a browser context).
  • Sanitization: Absent (no explicit validation or sanitization of text content before it is placed into the diagram JSON).
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 03:34 AM