Privacy

Off Your Shelf keeps your reading list on your device. This page says exactly what leaves it, and where it goes.

Who is responsible

Andreea Bahnareanu, Vienna, Austria, a private individual running this site as a non-commercial project (see the Impressum). Contact for anything in this notice: hello.offyourshelf@gmail.com.

What stays on your device

Your library, the books you mark “I’m reading this”, and your “Does this fit?” answers (each saved with the mood you searched) are stored in your browser’s local storage on this device. They are not stored on our servers. Marking a book “I’m reading this” also stops it being recommended again; it is reversible with the Undo shown right after you choose it.

When you import your Goodreads shelf

Your CSV file is read in your browser and is not uploaded. To add page counts, subjects and descriptions, your browser then looks each book up on Open Library, sending its ISBN, or its title and author when there is no ISBN. Covers for the books we recommend are loaded from Open Library too. Open Library sees these requests, including your IP address, like any website you visit.

When you ask for recommendations

Your mood and your books’ details (title, author, page count, subjects and description) are sent to our server, which passes them on to an AI model to pick your matches. The whole shelf goes every time, because the model has to see the list it is choosing from. Today the model is provided by OpenAI and reached through OpenRouter; if that changes, this page will change with it.

A mood can say more about you than you meant it to. Please keep it to how you want your reading to feel and leave out health or other sensitive personal details, because it goes to an AI provider.

We do not save your mood or your books, and our server logs contain no mood or book content. We ask OpenRouter to route only to providers that do not collect user data.

“See how it works” uses a fixed demo shelf of well-known books that lives on our server, so only the mood you type or the starter you pick is sent, never your library. Starter answers are normally pre-computed, in which case nothing is sent to an AI provider at all.

What the providers say about their own handling, as we last checked (23 September 2026):

• OpenRouter states that it does not log prompts or responses by default, keeps basic request metadata (timestamps, the model used, token counts), and does not use your inputs or outputs for model training.
• OpenAI states that data sent to its API is not used to train its models unless the customer opts in, and that its abuse-monitoring logs can be kept for up to 30 days.

Their own policies apply to the data they receive, and can change. We link to them rather than restate them.

Our host and abuse protection

The site runs on Netlify. Like any web host, it receives your IP address and request details when you use the site, and it uses IP addresses to rate-limit requests so the service can’t be abused.

Cookies and analytics

Off Your Shelf does not set cookies and currently runs no analytics or advertising.

Why this data is used, and on what basis

• Looking up your books on Open Library and sending your mood and shelf to an AI model, to give you the recommendations you asked for (Art. 6(1)(b) GDPR).
• Hosting the site and rate-limiting requests, to keep it available and protect it from abuse and runaway costs (legitimate interest, Art. 6(1)(f) GDPR).

You choose whether to import a shelf or ask for recommendations. Without that, the app has nothing to work with and sends nothing.

Who receives data, and transfers outside the EU

• Netlify (hosting), OpenRouter (routes the request) and OpenAI (the AI model) are based in the United States.
• Open Library, run by the Internet Archive in the United States, receives requests that your browser sends directly. They do not pass through us.
• Google (Gmail) handles the emails you send us, and is based in the United States.

Transfers to Netlify, OpenRouter and OpenAI rely on the European Commission’s standard contractual clauses; Netlify also states that it is certified under the EU–US Data Privacy Framework. Emails handled by Google rely on Google’s own certification under that framework. This is what the providers’ documents say, as we last checked (4 October 2026).

How long data is kept

We store nothing about you ourselves: no moods, no books, no account. Netlify keeps standard request logs under its own retention periods. OpenAI may keep abuse-monitoring logs for up to 30 days, as mentioned above. The analytics service keeps aggregate counts without IP addresses.

Your rights

You can ask for access to your data, correction, deletion, restriction or a copy of it, and you can object to processing based on legitimate interest. Because we hold almost nothing that identifies you, there is usually little to find; what your browser stores you can delete yourself (below). Write to hello.offyourshelf@gmail.com.

You also have the right to complain to the Austrian data protection authority (Datenschutzbehörde) or to the authority in your own country.

The recommendations are produced by an AI model, but nothing is decided about you with legal or similarly significant effect.

Deleting your data

Everything is stored on your device, and we hold no copy to delete. “Replace shelf” in the header replaces your library with a new Goodreads export once it has been read successfully, and resets which books are excluded from recommendations. It does not remove your saved choices and answers, so to remove everything, clear this site’s data in your browser settings.

Contact

Questions about privacy: hello.offyourshelf@gmail.com. If you write to us (here or on the contact page), we use your email address and message only to answer you, on the basis of our legitimate interest in replying (Art. 6(1)(f) GDPR), and keep them for as long as the conversation needs. Emails go through Google’s Gmail, which is based in the United States.

Last updated 4 October 2026.