Verify emails with an AI agent
Connect WarmySender to Claude, ChatGPT, Cursor, Codex, OpenClaw or Hermes Agent — or any agent that speaks MCP — and ask it in plain language to check a single address or a whole list. It reads the results back, turns the confirmed addresses into a list, and builds the campaign around them, in the same conversation. What it cannot do is send to them: your agent never sends a message itself and can never raise a limit.
That last part is the difference worth knowing. A verification tool that only verifies leaves you exporting a file and starting again somewhere else. Here, "check these 800 addresses and build me a campaign for the good ones" is one sentence, and the campaign that comes out is a normal WarmySender campaign, paced by the same scheduler as everything else.
What can my agent actually check?
| Ask for this | What happens |
|---|---|
| One address, right now | Your agent checks it and tells you whether it is safe to send to, risky, or not deliverable |
| A whole list | Your agent submits the batch and reports progress; batches run up to 5,000 addresses at a time |
| "How is that batch doing?" | It lists your batches and reads back the results of any of them |
| "Stop that one" | It cancels a batch that is still running |
| "How much have I got left?" | It reads your remaining verification allowance before you commit to a big list |
| "Make a list of the good ones" | It turns the confirmed addresses from a finished batch straight into a prospect list, ready to enroll |
| "Now email them" | It builds and starts the campaign — and WarmySender's scheduler, not your agent, sends every message at your normal safe pace |
| "Find me their email addresses" | Not something WarmySender does. Verification confirms addresses you already have; to get addresses you do not have, use the built-in business database or import a list that already contains them |
What does the verify-then-launch flow look like?
- Check what you have left. "How much verification allowance do I have this month?"
- Hand over the list. "Verify these 640 addresses." Your agent submits them and tells you when the batch is done. Big list? Split it — each batch runs up to 5,000 addresses.
- Keep the good ones. "Make a list called Q4 Prospects from the addresses that came back safe to send."
- Build the campaign. "Create a three-step campaign for that list, send from my two warmed mailboxes, weekdays only, and start it."
- Watch it. "How did that campaign do this week, and did anything bounce?"
Step 4 is the one people expect to be blocked and it is not — your agent really can start the campaign. Starting it simply hands it over: the scheduler then spreads the sends across your sending window, inside each mailbox's daily limit and your domain's ceiling. See how many cold emails you can send per day for the numbers.
How does my allowance get spent?
- Your agent shares your allowance — it does not get its own. Verifying through an agent draws on exactly the same monthly allowance and credits as verifying in the app.
- You are only charged for answers. One credit per confirmed result; an address we cannot confirm costs nothing.
- Credits never expire and top up your plan's monthly allowance whenever you need more.
- Your agent cannot go faster. Verification is paced for the same reason sending is, and an agent asking for a rush does not change the pace.
- It can check before it spends. Ask it to read your remaining allowance first — a habit worth forming before a big list.
What about leads from the built-in database?
Your agent can search the built-in business database, save what it finds and export it. Treat those addresses like any other list: verify them before you send. A record in a database is a record, not a promise about today's mailbox — people change jobs, addresses get retired, and a clean list is what keeps your bounce rate down. The pleasant version of that chore is one sentence: "Save these 200 leads, verify them, and build the campaign from the ones that come back clean."
What can't an agent do here?
- Send anything. Ever. It builds and launches; WarmySender sends.
- Raise a limit. Not your verification allowance, not a sending cap, not a ramp stage.
- Connect or disconnect an account. That stays with you in the app, so no credential passes through an agent.
- Find an address you do not have. Verification confirms; it does not discover.
- Promise a result. Some mailboxes cannot be confirmed either way — your agent will say so rather than guess, and those cost you nothing.
Frequently asked questions
Which AI agents can verify emails for me?
Claude, ChatGPT, Cursor, Codex, OpenClaw and Hermes Agent, plus any agent that speaks MCP. One WarmySender key works with all of them, so you can verify from whichever agent you already use. Setup is on the connect your AI agent page.
What can my agent actually check?
A single address on the spot, or a whole batch of up to 5,000 addresses. It can list your batches, read results back, cancel one that is still running, read your remaining allowance, and turn the confirmed addresses of a finished batch into a prospect list you can enroll straight away.
Can my agent send to the addresses that came back clean?
It can build and start the campaign for them — but it never sends the messages itself. Starting a campaign hands it to WarmySender's scheduler, which paces every email inside your sending window, your per-mailbox daily limit and your domain's ceiling. That is true whether a person or an agent pressed go.
Can my agent raise my verification allowance?
No. Your agent shares the same monthly allowance and credits you use in the app and cannot add to them or verify any faster than the normal pace. It can, however, read how much you have left before starting a big list, which is the useful half of that question.
How am I charged when an agent verifies?
Exactly as you are in the app: one credit per confirmed result, and nothing at all for an address that cannot be confirmed either way. Credits never expire, and they top up the monthly allowance included with your plan.
Can my agent find email addresses for my contacts?
No — verification confirms addresses you already have rather than discovering new ones. If you need addresses, the two customer paths are the built-in business database, which your agent can search and save from, and importing a list that already contains them. Once you have them, verifying is the sentence that comes next.
Are leads from the built-in database already verified?
Treat them as unverified. Database records are a starting point, not a statement about a mailbox today — people move on and addresses get retired. Ask your agent to verify a saved list before you send to it, and build the campaign from what comes back clean.
Related reading
Still have a question? Email [email protected] — we respond within 24 hours on business days.