excalidraw-skill

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources, creating an indirect prompt injection surface. Ingestion points: Diagram data from .excalidraw and JSON files via import scripts, and scene details from the describe_scene tool. Boundary markers: The skill relies on behavioral instructions like a 'Quality Gate' rather than technical delimiters. Capability inventory: Scripts and tools can read/write to the file system and perform network requests to the canvas API. Sanitization: No data validation or sanitization was observed for the ingested elements.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 02:30 PM
Security Audit — agent-trust-hub — excalidraw-skill