excalidraw-architecture

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted data by reading repository source files (e.g., package.json, Dockerfile, and source code) to identify architectural components. This creates a surface for indirect prompt injection where malicious instructions embedded in codebase comments or metadata could attempt to influence the agent's behavior.
  • Ingestion points: Codebase entry points and configuration files.
  • Boundary markers: Absent; instructions do not explicitly tell the agent to ignore instructions found within analyzed files.
  • Capability inventory: File system write access to the docs/ directory.
  • Sanitization: Absent; the content is parsed to generate a JSON diagram.
  • [EXTERNAL_DOWNLOADS]: The skill references excalidraw.com as the schema source and suggests the user visit the site to view diagrams. Excalidraw is a well-known service, and these references are legitimate and safe.
  • [COMMAND_EXECUTION]: The skill suggests a shell command (open docs/architecture.excalidraw) to the user as a helper to view the result on macOS. This is an informative suggestion for the user rather than an automated execution by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 08:22 AM
Security Audit — agent-trust-hub — excalidraw-architecture