mermaid-diagram-design

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes custom shell scripts (scripts/assess_mermaid_density.sh and scripts/validate_mermaid_markdown.sh) used for analyzing diagram complexity and validating syntax. These are intended to be executed by the agent as part of the diagram design workflow.
  • [EXTERNAL_DOWNLOADS]: The script scripts/validate_mermaid_markdown.sh utilizes npx -y @mermaid-js/mermaid-cli to download and execute the official Mermaid command-line tool. This is a standard practice for verifying that diagrams can be correctly rendered and follows the guidelines for utilizing well-known services.
  • [PROMPT_INJECTION]: The skill implements a validation workflow that processes external markdown files, creating a potential surface for indirect prompt injection.
  • Ingestion points: Markdown files passed as arguments to the validation scripts in the scripts/ directory.
  • Boundary markers: None explicitly present in the shell scripts, though they perform static analysis or pass content to an external renderer.
  • Capability inventory: The skill can read local files and execute shell commands (via its own scripts and npx).
  • Sanitization: Input file paths are quoted in shell commands to prevent basic argument injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 04:06 AM
Security Audit — agent-trust-hub — mermaid-diagram-design