The Agent Skills Directory

[DYNAMIC_EXECUTION]: The scripts/lib/browser.mjs utility dynamically resolves and imports the playwright or playwright-core library from the current working directory (process.cwd()). This allows the skill to leverage the host project's browser automation environment but involves importing code from paths computed at runtime.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external URLs to perform audits, creating a potential surface for indirect injection.
Ingestion points: scripts/lib/pages.mjs fetches HTML and XML sitemaps from external URLs. scripts/observe-runtime.mjs renders external sites and captures network traffic, cookies, and console logs.
Boundary markers: No specific delimiters or safety warnings are used to wrap the tool's findings when presented back to the agent.
Capability inventory: The skill executes local Node.js scripts that drive a headless browser and perform filesystem operations (screenshots).
Sanitization: The scripts/lib/html.mjs parser extracts text and structure for analysis but does not sanitize content against potential instructions targeting the LLM.

mobile-readiness