AI Assistant · the wedge

Stop reading statuses.
Start asking.

Every other verifier hands you a wall of codes to decode. Ours reads your list, answers in plain English, and does something about it. Upload, then just ask: “Is this safe to send?”

Confirm-first spendPlain-English answersNever trained on your data
ai assistant · list #4821
Is my latest list safe to send to?
⚠ Grade C — send with caution
✅ Valid23,636 · 77.5%
❌ Invalid2,444 · 8.0%
⚠ Catch-all192 · 0.6%
❓ Unknown4,230 · 13.9%
Export the 23,636 valid addresses and exclude the invalids. The 4,230 unknowns are likely greylisted — I can re-verify those free. Want me to?
Yes, and download the clean ones
Plain English
every verdict explained
Confirm-first
never spends without you
14 tools
one MCP server, your stack
24/7
it never sleeps
Talk to your data

Ask one question. Get a real answer.

No CSV gymnastics. The assistant reads the whole list, grades it, and tells you what to do — with the numbers right there to back it up.

C77.5% valid
Valid23,63677.5%
Unknown4,23013.9%
Invalid2,4448.0%
Catch-all1920.6%

Deliverability grade on list #4821 — graded from real verdicts, not a guess.

ai assistant · list #4821
Is my latest list safe to send to?
⚠ Grade C — send with caution
✅ Valid23,636 · 77.5%
❌ Invalid2,444 · 8.0%
⚠ Catch-all192 · 0.6%
❓ Unknown4,230 · 13.9%
Export the 23,636 valid addresses and exclude the invalids. The 4,230 unknowns are likely greylisted — I can re-verify those free. Want me to?
Yes, and download the clean ones
Interpretation, not labels

What a catch-all means for your send.

A status code is trivia. A recommendation is value. The assistant turns every verdict into a next step — and is honest when a result is uncertain instead of faking a green light. Not a green light. An answer.

  • Catch-all?“Acceptance can't be confirmed — send to a low-risk warmup segment.”
  • Typo?“Did you mean gmail.com? I can fix and re-verify it free.”
  • Unknown?“Likely greylisted — free to re-verify for a clearer result.”
[email protected]Catch-allWarmup segment only
[email protected]TypoFix → gmail.com, re-verify free
[email protected]UnknownGreylisted — re-verify free
[email protected]ValidSafe to send
Agents · MCP server

Put verification inside your agents.

The same tool layer that powers the in-app assistant is exposed over Model Context Protocol. Point Claude Desktop or a custom agent at one URL and it inherits all 14 tools — scoped by key, metered by credits, confirm-first on anything billable or destructive.

  • One config block — no SDK, no glue code, no per-tool wiring.
  • Scoped & metered — keys you can revoke, spend you can cap per agent.
  • Same guardrails— agents can't spend a credit past your ceiling without a yes.
json
// Claude Desktop · mcp.json
{
  "mcpServers": {
    "emaillistchecker": {
      "url": "https://platform.emaillistchecker.io/mcp",
      "headers": {
        "Authorization": "Bearer elc_live_•••••"
      }
    }
  }
}

// Your agent now has 14 scoped tools:
//   verify_email · grade_list · find_email
//   list_jobs · read_analytics · sync_integration …
// every billable call is metered by credits,
// every destructive call is confirm-first.
14 tools in the registry
verify_emailverify_batchgrade_listfilter_verdictsfind_emaillist_jobsjob_resultsread_analyticsexport_cleansync_integrationcheck_creditsestimate_costre_verify_unknownsinbox_placement
The guardrail

It always asks before it spends.

No silent charges, no surprise deletions. Every billable or destructive action is proposed with its exact credit cost and waits for your yes. Unknown verdicts stay free — you only pay for answers we're sure of, and credits never expire.

  • Exact cost up front — shown in credits, before anything runs.
  • One-click approve — or tweak the scope and ask again.
  • Fully logged — every agent action is auditable after the fact.
Confirm before I run this

Re-verify the 4,230 unknowns in list #4821 and export the clean set.

Re-verify unknowns (greylisted)0 credits · free
Export 23,636 valid addresses0 credits
Total0 credits
Approve & run
It does the work

An assistant that acts — safely.

Verify, clean, segment, export, sync. It proposes every billable or destructive action with its exact cost and waits for your yes.

Confirm-first by design

It never spends a credit or deletes a thing without showing you the cost and getting your approval first.

Agents + MCP

Expose EmailListChecker as a secure tool to your own AI agents — scoped by key, metered by credits, same guardrails.

Upload in chat

Attach a CSV or Excel file and it verifies it for you.

Reads your analytics

Ask why your valid rate dropped — it digs in.

Finds & verifies

“Find Jane's email at acme.com” — done.

Loved by senders

The part teams actually love.

We cut our bounce rate from 11.4% to 0.5% in a week. The AI explaining why each address was risky is the part our team actually loves.

Sarah ChenHead of Growth, TechFlow

The API drops straight into our signup flow. Fake and disposable emails never make it into the database anymore.

David KimEngineering Lead, DataSync

Half the price of the tool we left, and the inbox-placement testing paid for itself on the first campaign.

Emma ThompsonCRM Manager, GrowthLab
Questions

Good to know.

No. It operates the platform through the same tool layer the app itself uses — 14 real tools that verify emails, grade and filter your lists, read your analytics, manage integrations and check credits. The chat is the interface; the work is the product.

Stop decoding statuses. Just ask.

100 free verifications, no credit card. Talk to the AI that cleans your list in under a minute.