01
Prevent bad downstream actions
Check the address before the agent creates a contact, assigns an owner, starts a sequence, or sends an invitation.
Product / MCP for AI agents
A verification tool your AI can call
Give every agent a deliverability checkpoint.
Connect the MailTooth remote MCP server once and let compatible AI tools verify an address before they create a record, trigger a workflow, or send a message.
- One governed verify_email tool
- Structured evidence
- Quick, Standard + Deepcheck
Automation needs a truth check
An agent can move quickly—and amplify bad contact data just as quickly.
Without a verification step, automated research, enrichment, CRM updates, and outbound actions can turn one guessed address into wasted work across an entire workflow.
02
Replace prompt guesswork
Give the model an explicit tool and structured result instead of asking it to infer deliverability from the address alone.
03
Keep policy in your control
Define which verification depth to use and which valid, invalid, risky, or unknown outcomes require review.
Productivity that compounds
Add one reusable checkpoint to many agent workflows.
The same tool can protect the moments where an agent turns discovered information into action.
Lead-intake agents
Verify a discovered address before creating the CRM contact or routing it to a sales owner.
Outbound agents
Use Deepcheck for high-value recipients before drafting or triggering an approved send.
Signup and onboarding agents
Check an address before accepting a record, sending an invite, or starting an onboarding workflow.
Data-hygiene agents
Recheck stale records and route invalid, disposable, role-based, risky, and unknown outcomes differently.
Support operations
Validate an updated contact address before an automated system changes the customer record.
Research assistants
Confirm one selected prospect while preserving a human-approved boundary around larger lists and spend.
From intent to evidence
A native tool call inside the agent’s work.
No custom parsing layer is needed between the agent and the verification result.
- 01
Connect the remote server
Add the stateless Streamable HTTP MCP endpoint to a compatible client and authenticate with a MailTooth API key.
- 02
Let the agent call verify_email
The tool accepts one email and a chosen verification depth, then returns the same evidence used across the platform.
- 03
Route the next action
Use address, mailbox, risk, role-based, disposable, and credit signals to continue, stop, retry, or ask for review.
Built for governed automation
Faster agent work without a second verification stack.
MCP uses the same account controls and verification engine as the REST API.
Structured tool output
Receive concise content plus the complete structured verification result for reliable routing.
Three evidence depths
Use Quick for screening, Standard for domain intelligence, and Deepcheck for mailbox-level evidence.
Shared credit ledger
MCP calls use the same balance, usage history, refunds, cache, and quotas as API requests.
Scoped API keys
Create a governed key for the agent, monitor usage, and revoke access without changing the integration.
Broad client compatibility
Works with compatible Streamable HTTP clients and agent frameworks, including custom clients.
Non-destructive by design
The verification tool reads evidence and returns a result; your workflow remains responsible for the final action.
Your first 100 credits are free
Let your agents move faster on cleaner contact data.
Start without a card. Choose the workflow and verification depth that match the decision in front of you.