The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/check.mjs performs dynamic path calculation to locate Node.js modules. It attempts to resolve the global npm root via environment variables (process.env.PW_ROOT) or shell commands, then uses createRequire to load the playwright library from that computed location.
[COMMAND_EXECUTION]: The validation script scripts/check.mjs executes the shell command npm root -g using execSync to determine the directory for global packages during runtime.
[INDIRECT_PROMPT_INJECTION]: The skill acts as a data processing pipeline that ingests user-provided Markdown and renders it into HTML for rendering in a browser via Playwright.
Ingestion points: Markdown files (<tên>.md) defined in the input contract.
Boundary markers: Uses custom block markers (e.g., === page, === cover) to delimit content sections.
Capability inventory: scripts/book.py performs file system operations; scripts/check.mjs executes shell commands and manages browser sessions.
Sanitization: scripts/book.py employs html.escape to sanitize Markdown text before it is embedded into the HTML output.
[EXTERNAL_DOWNLOADS]: The conversion script scripts/book.py embeds links to Google Fonts (fonts.googleapis.com and fonts.gstatic.com) to provide the required typography for the document layout.

book-layout