aim-page-fetch
npm run page-fetch -- "https://example.com"
npm run page-fetch -- --screenshot /tmp/page.png --url "https://example.com"
Defaults: minimized, stop after session (--keep-open to chain).
JSON fields: title, url, canonical, description, meta, mainText, links[], rawLength, optional screenshot, optional cached.
Parse markdown, then ---JSON---. Prefer this over guessing page text.