Write to a wallet.Only that wallet reads it.

Ciphra seals every message in your browser with a key your wallet derives. The server keeps ciphertext it has no way to open.

Signing is free. No transaction, no fee. X25519 keys, XSalsa20-Poly1305 boxes, padded to fixed sizes
To the wallet you nameplaintext, still in your tab
Nonce…
Recipient key…
Stored size…
Server can readNothing

Scroll. This line is being sealed for real, right now, with a throwaway key made in this tab. Nothing leaves it.

Your wallet signs one fixed sentence. From that signature your browser derives a messaging key that never leaves it.

The public half goes into a directory, signed by your wallet, so anyone writing to you can check it is really yours.

Messages are sealed to that key before they are sent. What reaches the server is a box, a nonce and two public keys.

Write to a wallet that has never been here and the note waits, sealed to you, until they publish a key. Then it is sealed to them.

Your wallet signs once
Key in your tab never sent
Plaintext stays here
Box ciphertext
Nonce 24 bytes
Public keys signed

One row on the server

sender
your wallet
recipient
their wallet
box
base64 ciphertext
nonce
24 random bytes
expires
never, 1 h, 1 d or 7 d

The real counts are on the wall below.

Wallets with sealed mail they haven't opened

Counts only. Nobody, including us, can see what the messages say. A wallet can take itself off the wall from its inbox.

Reading the wall…

See the whole wall

Three reasons people write sealed

One key, one inbox, any wallet you can name.

Found a bug in a contract, or a leaked key in a repo? Tell the deployer wallet directly, without posting it in a public reply.

  • Write to the deployer address
  • Or its .sol name
  • Set it to burn after a day
  • They see it on next sign-in
Write to a dev

Ask before you sign

Short answers here, the whole protocol in the docs.

Pick a question. Every answer is also written out on the docs page, with the code that does it.