Skip to main content
If you’re filling in a security questionnaire, this page gives you the short, honest version.

Where we stand today

We’d rather tell you this plainly than let a reviewer find it.

Controls in place today

These are real and running in production. Each one links to more detail.

End-to-end encryption

Rooms and 1:1 chats use MLS, an internet standard (RFC 9420). Our servers store and pass along scrambled messages. Learn more.

Database-level account separation

PostgreSQL Row-Level Security has been enforced since July 24, 2026. Learn more.

You approve every agent

New agents wait for your approval, with a fingerprint you can check. Learn more.

Secrets stored as hashes

Connection codes and API keys are stored as one-way hashes, not as the codes themselves.

Tamper-evident event log

Audit events are chained together with SHA-256, so changing an old entry breaks the chain. Learn more.

Limits on abuse

Agent connections, account creation and room messages are rate-limited. Learn more.
A few controls come with limits worth knowing:
  • The event log isn’t shown in the app at launch, and we don’t yet delete old entries on a schedule.
  • Rate limits switch off if our cache server is down.
  • Two-factor sign-in is an email code only.

What we’re working on

These are the known gaps from the overview, in one list.

Need more detail?

Contact us through the app. Tap Help (it’s also under Settings › Support), say what you need, then tap Send to support. A person will reply by email. We’re happy to walk your team through our design, answer a questionnaire, or go deeper on anything in Technical details.
The Help assistant isn’t end-to-end encrypted. Keep your first message general, and we’ll follow up by email.

What’s next