Privacy

Privacy

Written so you can check it from the page source, rather than take our word for it.

There is no account here, no login, no newsletter, no email box, no comment field and no analytics script. Nothing you type on this site is sent anywhere — with one exception, the two AI tools, which have to send what you paste in order to answer it, and which keep none of it.

What is kept in this browser, key by key

Your work is kept in this browser. That phrasing is deliberate, and it is not a synonym for permanent — see the next section. This is the complete list of what can be written, and one file,src/lib/storage.ts, is the only place in the codebase allowed to write any of it.

tgr.theme.v1

The word “dark” or the word “light”.

Only when you press the theme button in the header. Never on page load.

tgr.log.v1

A list of the things you have ticked or typed on this site. Each entry is four fields: t, a timestamp in seconds; k, a kind (build, play, day, field or score); id, which thing it was; and v, the value — 0 or 1 for a tick, the text you typed for a worksheet field, a number for a score.

Only when you tick something or type something. Every write comes from a tap or a keypress you made.

That is the whole list. There is no device id, no visitor id, no session id, no "last page you were on", and no A/B bucket. If there were, this site would owe you a consent banner, and it does not have one.

You can read all of it and delete all of it fromyour receipts page, or by clearing site data in your browser. There is nothing held anywhere else to delete, because there is nowhere else.

“Kept in this browser” does not mean permanent

Browser storage is a cache, and we treat it as one. Safari deletes everything a site has written after seven days of Safari use in which you do not come back. School Chromebooks wipe local data when you sign out. Private browsing keeps nothing at all. A new phone starts empty.

So the durable record is the one you take with you: the printable receipt, and the copy-as-text button on every worksheet, which puts your answers in your own notes app in one tap. Use them. Losing what is in this browser should be annoying, never expensive.

Not because we found a loophole. The ePrivacy rules are about storing things on your device, whatever the technology is called — swapping cookies for localStorage does not escape them. The only exemption that applies is for storage that is strictly necessary for something you explicitly asked for.

Choosing a theme and ticking a build you have done are both things you explicitly asked for. So the rule the code enforces is: every write must originate in a gesture you made, and in development a write without one throws an error rather than shipping quietly.

What our host sees

We collect nothing. Our hosting provider processes connection metadata on our behalf, as our processor — that is what serving a web page means. When your browser asks for a page, the server sees your IP address and your browser's user-agent string. An IP address is personal data under the GDPR, so we are not going to pretend otherwise.

Host details pending

This section must name the hosting provider, say exactly what its access logs contain, state the retention period, and say whether logging is disabled or IP addresses are truncated. The host is not chosen yet. Publishing this page without those specifics would be the vague version of privacy this site exists to avoid.

The two AI tools

Scam Check andShrink My First Step are the only parts of this site that send anything off your device, and neither is live yet. When they are:

  • What you paste goes to an AI model, is answered, and is dropped.
  • Nothing is stored — not the input, not the output, not a log of either.
  • There is no account, so nothing can be attached to you in the first place.
  • Both tools tell you on the page that they are AI, and both say the same thing about storage that this page does.
  • Take names, addresses, phone numbers and links out before you paste. The tools do not need them and the answer is no better with them in.

The trade-off is stated openly in our decisions log: storing nothing also means we have no record if the tools are misused. We chose that deliberately.

Third parties, and why the page source proves it

There are no third-party requests on this site. The fonts are self-hosted files served from this domain, there is no CDN, no tag manager, no embedded video, no social widget and no tracking pixel. A step in our build reads every file that gets published and fails the release if any external host appears in it.

That is the part worth checking rather than trusting: open the network tab, load any page, and count the domains. There is one.

There is no analytics on this site today. If that ever changes it will be cookieless and aggregate — never anything that identifies a person — and it will be listed on this page and in the changelog before it ships.

AI crawlers are allowed in, on purpose

The robots file disallows nothing and blocks no crawler. It names OAI-SearchBot, GPTBot, ChatGPT-User, PerplexityBot, ClaudeBot, Claude-Web, Google-Extended, Applebot-Extended and CCBot as allowed, rather than leaving it implied. Teachers, parents and librarians increasingly find material by asking an assistant, and this site has nothing to protect from being read: it is a book's worth of free tools and safety text with no user data behind it.

Questions, corrections, and schools

If you are reviewing this site for a school, there is a version written for you atschools: privacy, where every claim can be verified from the page source in a couple of minutes.

Otherwise the mailbox at the bottom of this page is the route. It is read by an adult, it is the only inbound channel on the whole domain, and there is no form anywhere on this site that collects anything from a child.