external-diagram-mermaid-generate

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The helper script scripts/query-org-metadata.py uses subprocess.run to call the official Salesforce CLI (sf). This is used to retrieve object record counts and Org-Wide Default sharing settings to enrich generated diagrams with accurate architectural context. The command execution uses static argument lists and avoids shell interpolation, ensuring safe operation.
  • [COMMAND_EXECUTION]: The scripts/mermaid_preview.py utility utilizes subprocess.Popen to launch the preview server as a background process. This is a legitimate use case for a developer tool, enabling a live-reload experience in the browser while the user iterates on diagrams.
  • [EXTERNAL_DOWNLOADS]: The preview server's HTML interface references the mermaid.js library from the JSDelivr CDN. This is a well-known and trusted source for the library and is used strictly for client-side rendering of diagrams.
  • [SAFE]: Analysis of the skill's instructions and asset templates found no evidence of prompt injection, data exfiltration, or obfuscation. The file access patterns are restricted to user-defined diagram files and internal temporary files (PID and status info) required for the preview server's operation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 12:31 PM
Security Audit — agent-trust-hub — external-diagram-mermaid-generate