mermaid-diagrams
Fail
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: HIGHCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The
scripts/validate-mermaid.shscript is vulnerable to command injection. The$FILEvariable is directly expanded by the shell into a Node.js execution string (node -e), allowing a specially crafted filename to break out of the string literal and execute arbitrary JavaScript or shell commands.\n- [PRIVILEGE_ESCALATION]: Documentation inreferences/pdf-rendering-pipeline.mdsuggests using the--no-sandboxand--disable-setuid-sandboxflags with Puppeteer. Disabling the browser sandbox is a significant security risk that could allow malicious diagram content to exploit the browser engine and gain unauthorized access to the host system.\n- [DYNAMIC_EXECUTION]: The validation script usesnode -eto evaluate code constructed from unvalidated shell environment variables, creating a direct path for arbitrary code execution via string interpolation.
Recommendations
- AI detected serious security threats
Audit Metadata