Librito

Privacy

Last updated: 2026-04-23

Librito stores the minimum data required to synchronise your highlights, notes, and books between your device and the web. This document describes exactly what we store, for how long, and who can reach it. These commitments are enforced by code — the scheduled jobs and API-level constraints are part of the repository, not policy claims.

What we store

What we do not store

Book transfer retention

When you upload a book through the web app, Librito keeps a copy in temporary Storage just long enough to deliver it to your device.

Server access to book files

Server code never reads the bytes of an uploaded book. All transit uses signed URLs: the browser uploads directly to Storage, and your device downloads directly from Storage. The server only mediates bookkeeping rows.

Per-user isolation

All content rows are scoped by user ID and protected by row-level security. Librito has no sharing features.

Pauses during inactivity

Librito's Storage is hosted on Supabase. On our free tier, the database pauses after a period of inactivity. While paused, scheduled cleanup jobs also pause — they catch up when activity resumes. This does not change the retention targets above once the database is active.

Versioning

This document lives in the librito-io/web git repository. Any change to our privacy commitments goes through a pull request.