How a self-destruct message works
You write the message and get a link. The text is encrypted in your browser with AES-GCM-256 before it is sent, the key stays in the link after the #, and the server keeps only ciphertext. When the recipient opens the link and presses reveal, the message is shown once and the stored ciphertext is overwritten in the same operation. The link is spent, and every later visit meets a burned page with HTTP status 410.