For website owners
Our fetcher.
Family Rumpus is a family assistant. Now and then, when someone in a family asks it something, our server reads a public page or picture to answer them. This page says what that fetcher reads, how to tell it is really us, and how to keep it off your site.
It doesn't crawl
Every request is made because a person asked for something, and it reads only what that answer needs. It doesn't follow links, build an index, or come back on a schedule, and nothing it reads is used to train a model.
What it reads
- A product's page and its preview picture. When a parent compares things to buy, each listed product is shown with the picture its page names for previews (og:image and the like). We read the start of the page for that tag, then the picture. We keep a small copy, no larger than 400 pixels, for 30 days, show it only to that family and only beside a link to your page, and then delete it. We never add tracking or referral codes to the link.
- Pictures of a subject. For a family's research report or kitchen screen, found through Wikipedia and Openverse. A report keeps only openly licensed pictures, credited to their authors.
- A picture someone asks for. When a family asks Rumpus to find a picture, or to save one from a web page into their project, we fetch that picture.
How to tell it's us
It names itself in every request:
User-Agent: FamilyRumpus/1.0 (+https://familyrumpus.com/bot; [email protected])
A name can be copied, so each request is also signed with Web Bot Auth (HTTP Message Signatures). Its Signature-Agent is "https://api.familyrumpus.com", and our public keys are at https://api.familyrumpus.com/.well-known/http-message-signatures-directory. A request without a valid signature from one of those keys isn't from us.
How to keep it off your site
Before it reads a product page it checks your robots.txt and follows the rules for FamilyRumpus, or for * when there are none for it by name. To keep it off every page of your site:
User-agent: FamilyRumpus Disallow: /
We keep what your robots.txt says for up to a day, so a change can take a day to reach us. A picture a family asks to save is fetched without checking it, the way their own browser would. To be left out of that too, or to have us stop showing your pictures, including the copies we already keep, write to us and we will.
Talk to a person
Questions, a problem with our requests, or a request to be left out: [email protected]. A real person reads every message.