Emailvalidation.io
Email Marketingemailvalidation.io by everapi validates email addresses in real time. Agents check a single address for deliverability, syntax, disposable and catch-all status, and read the remaining quota before a batch.
What This Integration Enables
Two operations is a position, not a gap. emailvalidation.io by everapi does one thing at the doorway of a record and does it in detail: a single address, checked live, returning a state, the specific check that produced that state, and every underlying signal separately. Format validity, MX presence, the SMTP check, whether the mailbox is a role account, whether the domain is a free or disposable provider, a score, and a proposed correction when the domain looks like a typo of a common one. It is a gate you put in front of a form or a checkout, not a hygiene tool you run over a list of two hundred thousand records.
The detail is what makes it useful to an agent. Three of the reasons this API returns describe a failure to reach the mail server rather than anything about the address, and a flow that treats those as a verdict throws away real customers on the strength of somebody else's outage. FlowRunner agents branch on the deliverable flag for the clean cases, queue the unreachable ones for a retry, and hand the genuinely ambiguous ones to a person. That last part is the whole point of FlowRunner's human-in-the-loop model: the agent takes the volume, and the judgment calls stay with someone accountable for them.
Without FlowRunner
With FlowRunner
Use Case Scenarios
-
The signup that never becomes a bad record
A submission arrives from Typeform. Before anything is written, the agent calls Validate Email with the catch-all test enabled, because the catch-all field comes back empty rather than false when the check is skipped, and empty is not the same answer as no. A deliverable result flows straight into HubSpot as a complete record. A role account or a free-provider result still becomes a record, tagged for the routing rules that care. An undeliverable result is held out of the mailable population and returned to the form with a prompt, so the correction happens while the person is still on the page rather than three weeks later in a bounce report.
-
The checkout address that has to reach a real inbox
A payment completes in Stripe and the receipt is about to go out. The agent validates the billing address first. When the state comes back with a reason describing an unreachable server, the flow does not mark the customer as invalid: it queues a retry and sends the receipt anyway, because a timeout at the recipient's host says nothing about whether the mailbox exists. When the state is genuinely undeliverable, the agent raises it to the support queue with the order attached, because a customer who has just paid and cannot be reached is worth a person's attention rather than a silent bounce.
-
Batch work that prices itself first
Before running a list of addresses through the API, the agent calls Get Quota Status, which is exempt from the quota and the rate limit and therefore free to poll. It compares the monthly and grace allowances against the size of the run and posts the result to Slack: how many addresses, how much of the allowance that consumes, and how much would be left. If the run does not fit, the flow splits it across periods rather than discovering the limit halfway through and leaving half a list in an unknown state.
Human-in-Loop Highlight
This API will sometimes tell you the address is wrong and suggest what it should have been. When a domain looks like a typo of a common one, the response carries a proposed correction alongside the verdict. It is a genuinely useful signal and it is also the one thing on this connector that an agent must never act on by itself. Correcting an obvious misspelling of a large consumer provider is safe. Correcting an unusual but entirely real company domain to the common one it resembles sends the invoice, the contract, or the password reset to a stranger at a different organization, and nothing in the response distinguishes those two cases. So the agent collects them instead of applying them. It posts the batch: "14 addresses came back undeliverable with a suggested correction. 11 are consumer providers with a one-character difference. 3 are corporate domains where the suggestion changes the company, not the spelling. Apply the 11, review the 3?" The reviewer answers in one message, the agent applies the accepted corrections across the whole set, and no customer receives mail meant for someone else because a score looked convincing.
Agent Capabilities
2 actionsValidation
1- Validate Email Checks one address live and returns the full picture: a deliverability state, the specific check that produced it, and each underlying signal separately, including format validity, MX presence, the SMTP result, role and free and disposable flags, a score, and a proposed correction when the domain resembles a common one. The catch-all test is a separate opt-in check, and leaving it off returns an empty catch-all field rather than a negative answer.
Account
1- Get Quota Status Reads the monthly and grace allowances with their totals, usage, and remainder. Exempt from both the quota and the rate limit, so a flow can poll it freely to price a run before deciding whether to spend anything on it.
Frequently Asked Questions
What can FlowRunner do with Emailvalidation.io?
FlowRunner agents can run Validate Email and Get Quota Status in Emailvalidation.io.
Does connecting Emailvalidation.io to FlowRunner require OAuth?
No. Emailvalidation.io connects to FlowRunner with an API key, no OAuth flow required.
Can Emailvalidation.io trigger a FlowRunner workflow automatically?
Emailvalidation.io doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.
Start building with Emailvalidation.io
Free plan, no card required. Connect in minutes.