The failure worth guarding against is an invented reason written onto a contact record, because whoever opens that contact in six months reads it as fact. The instructions require the property to read unknown when nothing explains a bounce or a departure. OpenCode holds that as well as anything does, and being model agnostic means you can move up the moment you catch it filling a gap.
Run a staged conversation — no account needed. The agent handles it for real while a simulated world answers its tool calls; nothing touches real accounts, and nothing is actually sent.
Hard bounce
550 5.1.1 recipient address rejected, user unknown. This is the only billing contact on the largest annual plan on the account, 18,400 USD renewing on 12 November, and two other addresses at the same domain are still receiving product mail normally. Loops shows no suppression recorded against it before today.
Using this template drops you into a guided setup. It asks exactly this, nothing else:
Connect Loops
One sign-in. The agent acts through your account, scoped to what this template uses.
Connect Slack
One sign-in. The agent acts through your account, scoped to what this template uses.
Audience rules
What each kind of event means for your audience and what should follow: which contact property records a bounce, an unsubscribe or a signup, which domains and plans are worth a Slack ping, and how many departures off one send count as a run.
Runs on OpenCode
Preselected for this page — connect your OpenCode account during setup, or switch to NoClick's built-in models with one click.
Watch it handle a test run
A staged conversation against a simulated world — then it’s live.
Where neither the payload nor the record explains what happened, the property says so. An empty field is recoverable and a confident wrong one is not.
Leaving a single mailing list is checked against the lists a contact still belongs to, and nobody is ever moved onto, off, or between them.
Open a handful of contacts a week after switching it on and read the properties beside the events that produced them. The tell is a reason that sounds too tidy, such as a bounce explained by something the payload could not have carried. Properties are editable, so a bad week is a tidy up rather than a disaster, but the fifteen minutes early is worth spending.
Free to start. Guided setup, a test run against staged conversations, and it's live.