beautiful-mermaid

Warn

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The scripts/render.ts script uses child_process.execSync to dynamically install the beautiful-mermaid package using npm or bun if it is not found in the environment.\n- [EXTERNAL_DOWNLOADS]: The skill fetches the beautiful-mermaid package and its dependencies from the NPM registry at runtime to support diagram rendering.\n- [PROMPT_INJECTION]: The skill workflow passes agent-generated Mermaid code into shell command arguments (--code). This creates a surface for command injection if the input is not correctly escaped.\n
  • Ingestion points: SKILL.md (Step 1: Generate or Validate Mermaid Code).\n
  • Boundary markers: None present.\n
  • Capability inventory: scripts/render.ts (execSync, file write), scripts/create-html.ts (file write), agent-browser (browser automation).\n
  • Sanitization: None detected in render.ts or create-html.ts.\n- [REMOTE_CODE_EXECUTION]: The scripts/create-html.ts script embeds SVG content directly into an HTML wrapper, which is then opened via agent-browser. This creates a potential XSS or code execution vector if malicious script content is embedded within a Mermaid diagram.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 19, 2026, 11:39 PM
Security Audit — agent-trust-hub — beautiful-mermaid