Security
How we protect your account
What the platform actually does, in plain terms. For how sessions, disputes and reviews are kept fair, see trust and safety.
Accounts and sign-in
- Passwords are stored as bcrypt hashes, never in a form that can be reversed. New passwords need at least 10 characters and are refused if they appear on a list of commonly breached passwords or contain your name or email.
- Eight failed sign-in attempts lock an account for fifteen minutes. Wrong passwords and wrong two-step codes count toward the same limit, so neither can be guessed around the other.
- Two-step sign-in is optional and available to everyone in Settings: a code from an authenticator app (the TOTP standard, RFC 6238) is asked for after your password or Google sign-in. Each code works once. Backup codes are stored only as keyed hashes and each works once.
- Sign-in with Google uses PKCE and a one-time state check, and links to an existing account only when Google says the email address is verified.
- Settings lists every device signed in to your account, and lets you sign any of them out, or all of them at once.
Sessions
- Sessions live in httpOnly cookies that page scripts cannot read. The access token is short-lived (fifteen minutes by default); the refresh token that renews it is stored only as a hash.
- Refresh tokens rotate on every use. If an old one is ever presented again — the sign that a cookie was copied — every session on that account is ended.
- Changing or resetting your password, or signing out everywhere, ends every existing session at once.
Every request
- Every request that changes something must carry a CSRF token that matches a cookie only our own pages can read (the double-submit pattern), compared in constant time. Only our own web origins may call the API from a browser.
- Rate limits apply per surface — sign-in, password reset, two-step codes, writes, uploads, starting new conversations — and overall. Most are counted per account once you are signed in, and per network address otherwise, so one busy school network does not throttle everyone on it.
- API endpoints validate their input — bodies, query strings, route parameters — against a schema before acting on it; anything that does not fit is refused.
- Our pages are served with a Content Security Policy, are refused framing by other sites, ask browsers not to guess content types, and require HTTPS in production.
Uploads
- Only an allowlist of file types is accepted, with size limits. Images, PDFs and archives must also start with the bytes their format requires, so a program renamed to look like a picture is refused.
- Files are stored under names we generate — nothing you upload decides where it is written — and are served with headers that stop browsers guessing their type.
- Notes sold on Entueri are never handed out as files: pages are turned into images and served one at a time, each request checking that the reader has access.
Data at rest and payments
- Secrets the database has to hold in a recoverable form are encrypted with AES-256-GCM: two-step sign-in keys, tutors’ payout destinations (bank account or UPI id) and tax identifiers. Payout destinations and tax ids have no plain-text column at all — only the encrypted value, a masked form for display, and a keyed fingerprint used to spot two accounts paid into the same place.
- Card details are entered with our payment provider and never reach Entueri’s servers.
- Messages and other public text are checked automatically for contact and payment details. What is checked, and what happens when something is found, is set out on our legal page.
Audit log
- Every decision staff make about someone — verifying a tutor, payouts and credit adjustments, conduct cases, disputes, review moderation, background checks, moving a student to a new tutor — is written to an audit log with who made it, when, and from which address.
- Opening the evidence for a dispute is logged too, because it shows staff a session two people had in private.
Reporting a vulnerability
If you think you have found a security problem, email security@entueriinc.com with what you found, the steps to reproduce it, and what an attacker could do with it. We will reply to confirm we have it and keep you told of the fix.
Please test only against your own accounts, do not access or change anyone else’s data, do not degrade the service for others, and give us reasonable time to fix a problem before you disclose it. We will not pursue anyone who reports in good faith within these limits. We do not run a paid bug bounty at the moment.
Machine-readable contact details are at /.well-known/security.txt.