The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/render.mjs uses execSync to execute the shell command npm root -g. This is used to programmatically determine the global Node.js package installation directory.\n- [DYNAMIC_EXECUTION]: The scripts/render.mjs script performs dynamic module loading by using import() on file paths constructed at runtime. This logic is used to locate the Playwright library across multiple potential system paths.\n- [INDIRECT_PROMPT_INJECTION]: The skill contains an attack surface for indirect injection as it processes SVG data using Node.js scripts and browser automation.\n
Ingestion points: SVG files are read as text by scripts/lint.mjs and scripts/render.mjs.\n
Boundary markers: scripts/lint.mjs includes regex patterns designed to detect and fail on SVGs containing <script>, <image>, external href, or url() references.\n
Capability inventory: The toolkit includes file system access (fs), shell command execution (execSync), and browser automation (Playwright).\n
Sanitization: The scripts/lint.mjs utility acts as a security filter for SVG content generated or handled by the agent.

illustration-framed-panel