Questions · Published 2026-09-29 · Updated 2026-09-29 · AgentiSend
Is AgentiSend safe to use in production?
AgentiSend is safe to use in production for the controls that are on: a budget on every key, a kill switch, an approval hold, and a public status page. It does not yet have a reviewed legal opinion or a mailbox product.
AgentiSend is safe to use in production for sending, with the limits written on the key. A new key starts with a budget. PATCH /limits/keys/{id} is how a person sets it, and the key cannot raise it. POST /limits/keys/{id}/kill pauses one key. POST /emails is refused when the budget is spent, the key is paused, or a hold is waiting. That is the control plane. It is not a promise about inbox placement, and it is not a certification.
Where to read the operator and the status
About says what the product is and how to write to hello@agentisend.com. Security says what is implemented and what is not, and where to report a problem. Status is the measured feed. The enforcement policy is the ladder: warn, throttle, pause, and the appeal. GET /trust/standing returns the account's reading against that ladder. GET /status is the same feed as the status page, as JSON.
The terms, the privacy page, and the security page say where review by counsel stands. This page does not upgrade that sentence.
What production sending still requires
Verify a domain with POST /domains and POST /domains/{id}/verify before you send as that domain. Until then, onboarding@agentisend.com delivers only to member sign-in addresses, and simulator addresses reach nobody. GET /domains/{id} shows which records resolved.
A send can still land in a spam folder after the receiving server accepts it. Why are my emails going to spam? is diagnosis, not a placement rate.
Next
Written by the AgentiSend team.