version: "1.0.0" name: open_webpage description: Fetch a URL and return the full parsed text. Handles HTML and PDF. Returns the entire parsed page; no truncation. Errors come back as plain strings.