> When you first join a chat room, a personal keypair will be generated and stored in browser storage. If you are the creator of this room, the client will also

> When you first join a chat room, a personal keypair will be generated and stored in browser storage. If you are the creator of this room, the client will also generate a keypair for the room I'm trying to wrap my head around the security model here. What bothers me is the key distribution problem and the fact that a server still has to handle key distribution messages. With PGP, a "web of trust" is assumed where people have static long-lived public keys you can verify using existing trust relationships. The key pair in this app is being generated as you join or participate which makes "web of trust" impossible to use. So,I am assuming the security model is similar to end to end chat apps like signal? Those protocols assume the initial key exchange was authenticated by the users out of band and allow users to verify the "safety number". They assume the users already know each other and can succesfully perform out of band authentication for at least the first key exchange. My question is, is there a "safety number" for mentant? Is it straight forward and obvious when navigating the UI? (Signal's isn't) and will the compromise of an old key pair (personal or room forward secrecy) result in compromise of future messages? If you security is done right I love the idea,especially if you can integrate existing user directories (AD/ldap for example) where user keys for the chat can be verified in the directory db.