Remote MCP server

Email verification your AI agent can use.

Connect BounceHound to Claude, Cursor or any MCP client and verifying a lead list becomes a sentence, not a workflow. No exports, no uploads, no scripts.

Get a free API key See setup

Claude Code
Claude Desktop
Cursor
ChatGPT
VS Code
Windsurf
n8n

Example conversation

You: Before you send the outreach emails, verify every recipient and skip anything risky.
Agent:
→ verify_emails (240 emails)
Done. 212 valid, 15 invalid, 9 catch-all, 4 disposable. I'll send to the 212 valid addresses only. Also, mike@acme.con looks like a typo for acme.com. Want me to fix it?

Setup

Connected in under a minute

Step 1
Create an API key
Sign up free (250 credits included) and create a key on the API & MCP page of your dashboard.
Step 2
Add the server
Paste one command or one JSON block into your AI tool. It's a remote server, so there's nothing to install or host.
Step 3
Just ask
“Clean this list before I import it.” Your agent picks the right tool, verifies the emails and reports back.

Configuration

Pick your client and paste

Replace YOUR_KEY with the API key from your dashboard.

MCP endpointhttps://bouncehound.com/api/mcpStreamable HTTP · Bearer auth

Run this once in your terminal.

terminal

claude mcp add --transport http bouncehound https://bouncehound.com/api/mcp \
  --header "Authorization: Bearer YOUR_KEY"

Add BounceHound to ~/.cursor/mcp.json (or Settings → MCP).

mcp.json

// ~/.cursor/mcp.json
{
  "mcpServers": {
    "bouncehound": {
      "url": "https://bouncehound.com/api/mcp",
      "headers": { "Authorization": "Bearer YOUR_KEY" }
    }
  }
}

Paste into claude_desktop_config.json and restart the app.

claude_desktop_config.json

// claude_desktop_config.json
{
  "mcpServers": {
    "bouncehound": {
      "command": "npx",
      "args": [
        "mcp-remote", "https://bouncehound.com/api/mcp",
        "--header", "Authorization: Bearer YOUR_KEY"
      ]
    }
  }
}

Add to .vscode/mcp.json in your workspace, then start the server from the MCP view.

.vscode/mcp.json

// .vscode/mcp.json
{
  "servers": {
    "bouncehound": {
      "type": "http",
      "url": "https://bouncehound.com/api/mcp",
      "headers": { "Authorization": "Bearer YOUR_KEY" }
    }
  }
}

Tools

Five tools your agent can call

Each tool returns a readable summary plus structured JSON, so agents can act on the results.

MCP tools, parameters and cost
Tool	Parameters	What it does	Cost
verify_email	email: string	Verify one address: status, reason, role/free/disposable flags and typo suggestion.	1 credit*
verify_emails	emails: string[] (max 1,000)	Verify a batch. Returns a status summary plus per-email results. Duplicates are skipped.	1 credit / email*
clean_email_list	emails: string[] (max 10,000) include_risky: boolean (optional, default false)	Returns only deliverable emails, with counts of removed addresses by status and reason.	1 credit / email*
suggest_email_fix	email: string	Local syntax + typo check, e.g. jane@gmial.com → jane@gmail.com.	Free
get_credit_balance	None	Returns the account's remaining credits.	Free

* Free when the address fails local checks (bad syntax, no mail server, disposable domain) or the result is “unknown”.

Try it

Things to ask your agent

Once connected, plain English is enough. The agent chooses the tool.

“Verify jane@acme.com”
→ verify_email
“Clean the emails in leads.csv and give me only the deliverable ones”
→ clean_email_list
“How many credits do I have?”
→ get_credit_balance
“Check these signups for typos and fake addresses, then tell me which to email”
→ suggest_email_fix · verify_emails
“Remove disposable and catch-all emails from this list before I import it into HubSpot”
→ clean_email_list
“Before you send the outreach emails, verify every recipient and skip anything risky”
→ verify_emails

Prefer plain HTTP? Every MCP tool is also available in the REST API.

FAQ

MCP questions

What is an MCP server?

The Model Context Protocol (MCP) is an open standard that lets AI assistants call external tools. BounceHound's MCP server gives your agent email verification tools it can use whenever a task involves email addresses.

Which AI tools work with it?

Any client that supports remote MCP servers over Streamable HTTP: Claude Code, Claude Desktop, Cursor, VS Code, Windsurf, ChatGPT connectors, n8n and more. Clients that only run local servers can connect through mcp-remote.

How much does it cost?

The same as the REST API: one credit per verified address. Unknown results, bad syntax, domains with no mail server and known disposable domains are free. suggest_email_fix and get_credit_balance never cost anything.

Is my API key safe in the config?

Your key is sent only to BounceHound over HTTPS. Treat it like a password: keep it out of shared repos, and revoke it from the dashboard if it leaks. You can create a separate key for each tool.

Can my agent verify a whole CSV?

Yes. Ask it to read the file and call clean_email_list (up to 10,000 emails per call). It gets back only the deliverable addresses plus counts of what was removed and why.

Put the hound on your list.

Get 250 free credits, verify your first list in minutes, and connect your AI agent whenever you're ready.

Get 250 free credits View pricing