The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill launches a Google Chrome or Chromium process via the spawn method to render web content. It manages the browser lifecycle using the Chrome DevTools Protocol (CDP).
[DATA_EXPOSURE_EXFILTRATION]: The skill performs network operations to fetch content from arbitrary, non-whitelisted domains based on user or agent input.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the web which could contain natural language instructions designed to manipulate the agent's behavior.
Ingestion points: The scripts/main.ts file takes a URL as input and fetches its full rendered content.
Boundary markers: The resulting markdown output does not include boundary markers or explicit instructions for the agent to ignore potentially malicious content within the markdown body.
Capability inventory: The skill has the capability to write files to the local disk and execute JavaScript within the headless browser context.
Sanitization: The skill utilizes the @mozilla/readability and turndown libraries to sanitize HTML content, effectively stripping scripts, styles, and other non-textual elements, though this does not mitigate natural language instruction injection.

baoyu-url-to-markdown