pretty-mermaid
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- EXTERNAL_DOWNLOADS (MEDIUM): The README.md states that 'Dependencies will be auto-installed on first run.' This indicates that the skill's scripts likely invoke a package manager (e.g., npm) to download code dynamically from the internet during execution. As the author 'imxv' is not a trusted source, this behavior presents a risk of supply chain or unverified code execution.
- COMMAND_EXECUTION (LOW): The skill is designed to be executed via CLI commands (e.g.,
node scripts/render.mjs). While this is the intended functionality, the lack of the actual script source code in the provided file list makes it impossible to verify if these scripts perform safe argument parsing or if they contain malicious subprocess calls. - INDIRECT_PROMPT_INJECTION (LOW):
- Ingestion points: The skill accepts external Mermaid diagram files (
.mmd) as input for rendering. - Boundary markers: No specific boundary markers or 'ignore' instructions are present in the documentation to prevent the model from following instructions embedded within processed diagrams.
- Capability inventory: The tool invokes an external library (
beautiful-mermaid) to generate SVG/ASCII output. There is a potential risk of SVG-based XSS if the underlying library does not properly sanitize the diagram content. - Sanitization: Sanitization logic cannot be verified as the execution scripts are missing from the analysis.
Audit Metadata