Client security
Where private keys live on your device, how the local message store is protected, and which platform features Keylane relies on.
Privacy & security
Keylane makes specific, bounded claims about what it protects. This section explains each one, where the protection comes from, and where it stops.
Security in a messenger is not one property. It is a stack of separate guarantees, each with its own assumptions, and each capable of failing independently of the others. A messenger can have flawless transport encryption and still leak everything through metadata, or hold no server-side data at all and still expose your conversations through an unlocked phone.
Rather than compress that into a single claim, we document it in four parts.
Where private keys live on your device, how the local message store is protected, and which platform features Keylane relies on.
What the relay stores, what it cannot see, how long anything survives, and what a full server compromise would actually yield.
The adversaries Keylane is designed to resist, and the ones it is not. Read this before deciding whether Keylane fits your situation.
The protocol specification: identity, authentication, key exchange, payload encryption, routing, attachments, and push.
If you read nothing else in this section:
Keylane does not protect against an adversary who controls your unlocked device, and it does not hide the fact that you are using Keylane from a network observer who can see your connection to the relay. It cannot recover your identity if you lose every device holding your keys, because it does not hold them either.
Those are design consequences, not oversights, and the threat model covers them properly.
Found something wrong in this documentation, or a vulnerability in the product? Security issues go to our disclosure page; everything else to support.
The engineering documentation above describes how the system behaves. The legal documents describe what we commit to: