Copy a website
Copy an existing live website, its pages as visitors see them plus its pictures, styles and fonts, into the files a Trillion public app is published from
Opens each page of the site in a browser in the cloud, waits for it to finish drawing, and keeps it as a visitor sees it with its scripts taken out; follows the site's own links up to the page limit and fetches same-site pictures, styles and fonts. Copied addresses point at the copy; omitted pages and files keep their original addresses. Returns site.json in your files: { files: { path: contents } }, which publish_public_app_definition takes as its files, and a list of what could not be copied and why. Works on public HTTPS sites; nothing needs installing.
What it asks for
- url (Needed, web address): The site's first page, like https://example.com/
- max_pages (Optional, number): The most pages to copy (default 60)
- paths (Optional, list): Pages to copy that no link reaches, like ["/pricing"]
- exclude_paths (Optional, list): Paths already copied in an earlier batch; they are not copied again
- width (Optional, number): Browser width in pixels the pages are drawn at (default 1280)
Use it from your AI
Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this tool by what it does and runs it.
Say to your AI: “Use Trillion to copy an existing live website, its pages as visitors see them plus its pictures, styles and fonts, into the files a Trillion public app is published from”
What your AI calls: run_tool {"tool_id":"tool:copy-a-site"}