Only hashes are stored, so these values are shown once and cannot be retrieved
later. Ask again with the same email and you get a fresh key.
Why we ask for a name and an email
Not for a mailing list. Two reasons, both practical: An incident notice needs somewhere to go. If a test rail breaks or a sandbox behaviour changes under you, the email is how you hear about it. It tells a real attempt from a script. Nothing here is verified, and we are not pretending otherwise. It is a speed bump, not a gate.Asking twice
Ask again with the same email and you get a new key for the same application. That matters more than it sounds: your funding intents, ledger entries, and webhook history stay where they are. Losing a key does not mean losing your work. Older keys keep working. Nothing revokes a credential something may still be running against. If you want one gone, revoke it explicitly. One email may collect ten keys a day. Past that you get a429 with
retryable: true, which is the API telling you to wait rather than to give up.
What a sandbox key can and cannot do
The application is created in the test environment and stays there. A live key
cannot touch a test rail and a test key cannot touch a live one, enforced
server-side rather than documented as a warning.

