mermaid-config

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The Health Check (Step 2) explicitly runs npm install (--prefix "$PLUGIN_DIR/scripts") to fetch packages from the public npm registry and Step 3 executes the plugin's Node scripts (render.js) from $PLUGIN_DIR/scripts, so untrusted third‑party code can be installed and run as part of the workflow.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The health-check runs npm install --prefix "$PLUGIN_DIR/scripts" at runtime which will fetch and install packages (from the npm registry, e.g. https://registry.npmjs.org) and thus downloads and can execute remote code (the beautiful-mermaid dependency) that the skill auto-installs and relies on.