Trust
Security, in plain words.
What protects your account, your characters and your renders today. When something here changes, the changelog says so.
How we protect it.
Accounts
Passwords are hashed with Argon2; sign-in codes are stored hashed and expire in 10 minutes. Five wrong passwords lock the account for a while.
Your data
Every account is isolated; another account's ids answer 404. Emails and IP addresses in logs and limits are keyed hashes.
Backups
Encrypted nightly backups with daily, weekly and monthly copies, and an automated restore drill that proves they work.
The web app
A content security policy for every kind of page, strict security headers, and media served with a type allow-list.
AI-specific defences.
Briefs are data, not instructions
Text you write is normalized and scanned for injected instructions before the Yarnix planner sees it; suspicious briefs get a stricter review.
Moderation on every step
Briefs, plans and every prompt sent to an image, video or voice model are checked; kids series use stricter rules.
No surprise spend
Every render has a quote and a money cap; nothing spends beyond the hold you approved.
Licensed building blocks
Every library, font and model is checked for a license that allows commercial use before it ships.
Found a problem?
Write to support@yarnix.cc with "Security" in the subject. Please give us time to fix an issue before you share it; we reply to every report.