Email verification has always been something you bolt onto a form, a script, or a CRM workflow. Today we’re launching something different: RocketVerifier is now an agent-native product. Any AI assistant that speaks the Model Context Protocol — Claude, ChatGPT, Cursor, and a fast-growing list of others — can verify emails, run bulk jobs, and check deliverability signals directly inside a conversation.

Why MCP, and why now

Model Context Protocol is quickly becoming the standard way AI assistants call external tools. Instead of writing a custom plugin for every assistant, you implement MCP once and every compatible client can use it. For an API-first product like RocketVerifier, that’s a natural fit: our entire surface area is already a clean REST API, and MCP is essentially a typed, agent-friendly wrapper around exactly that.

None of the major email verification providers — ZeroBounce, NeverBounce, Bouncer — have shipped anything like this yet. We think that’s a mistake, because the way people reach for tools is changing fast. Increasingly, “verify this email” or “clean this list” is a sentence typed into Claude or Cursor, not a trip to a dashboard.

Two ways to connect

We shipped both flavors of MCP server so you can pick whatever fits your workflow:

1. Local, via npm

npx @rocketverifier/mcp

Add it to Claude Desktop or Cursor with a single JSON snippet:

{
  "mcpServers": {
    "rocketverifier": {
      "command": "npx",
      "args": ["-y", "@rocketverifier/mcp"],
      "env": {
        "ROCKETVERIFIER_API_KEY": "rv_live_your_api_key"
      }
    }
  }
}

2. Hosted, zero install

If your client supports remote MCP servers over Streamable HTTP — like ChatGPT’s connectors — skip the install entirely:

URL:  https://api.rocketverifier.com/mcp
Auth: Bearer rv_live_your_api_key

Both expose the same ten tools, so switching between them is transparent.

What you can do with it

  • verify_email — real-time deliverability check with a 0-100 confidence score
  • create_bulk_job, get_job_status, get_job_results, list_jobs, cancel_job — the full bulk verification lifecycle
  • check_credits — check your balance without leaving the chat
  • check_disposable, mx_lookup, check_domain_auth — the same free tools available on our tools hub, free of charge inside any MCP client

In practice this means you can paste a messy list of emails into Claude and ask it to verify them, flag the risky ones, and summarize the results — no dashboard, no CSV upload, no context switch.

Try it sandboxed, no credits required

Every tool works with a sandbox key (rv_test_...) against @sandbox.rocketverifier.com addresses, so you can wire up the integration and see real tool-call behavior before touching production credits.

Get started

Read the full setup guide for Claude Desktop, Cursor, and ChatGPT in our MCP documentation, or jump straight to the MCP Server product page for an overview. If you want a hands-on walkthrough, we also published a tutorial on cleaning an email list with AI agents.

This is the first of several agent-native features we’re shipping — stay tuned.