EngageLab Agent Email MCP Server
Enables AI agents to send, receive, and reply to email through dedicated thread-aware mailboxes, with sandbox testing and compatibility across MCP clients.
README
EngageLab Agent Email MCP Server
English | 简体中文
Send, receive, and reply to email from AI agents — through the Model Context Protocol.
Each agent gets its own mailbox: a real address that sends and receives, with conversations grouped into threads so your agent has context. Works with Claude Desktop, Claude Code, Cursor, Cowork, and any MCP-compatible client.
{
"mcpServers": {
"engagelab-email": {
"command": "npx",
"args": ["-y", "@engagelabemail/mcp"],
"env": { "ENGAGELAB_EMAIL_SECRET_KEY": "sk_sg_your_key" }
}
}
}
Why
Most email APIs are send-only. Agents that need to have a conversation over email — support, scheduling, approvals — need an inbox.
- Send & receive — the agent owns a dedicated mailbox, not a bolted-on parse webhook
- Thread-aware — replies group into conversations, so the agent sees full history
- Sandbox mode — test agent flows without emailing real humans
- One env var — no OAuth dance for local use; region inferred from the key
Tools
| Tool | What it does | |
|---|---|---|
| Discover | list_mailboxes |
List mailboxes; call this first to get IDs |
| Send | send_email |
New email (text/html, cc/bcc, attachments, sandbox) |
reply_email |
Reply to a message — recipients & thread inferred | |
| Receive | list_inbound_messages |
Inbound mail, filter by mailbox/keyword |
get_message |
Full message details by UID | |
check_new_messages |
One-shot poll for new mail | |
| Threads | list_threads / get_thread / list_thread_messages |
Browse conversations |
Quickstart
1. Get a Secret Key. Create an EngageLab account and generate a key from the console (format sk_sg_xxx — the prefix selects the region). Or use the CLI: engagelab-email-cli login creates one via browser.
2. Create a mailbox. In the console, create a mailbox for your agent (shared subdomain is instant; custom domains need DNS verification). Programmatic mailbox creation is on the roadmap — see Troubleshooting if list_mailboxes returns empty.
3. Register the server.
Claude Desktop — add the JSON block above to claude_desktop_config.json.
Claude Code:
claude mcp add engagelab-email \
-e ENGAGELAB_EMAIL_SECRET_KEY=sk_sg_your_key \
-- npx -y @engagelabemail/mcp
Then ask your agent:
List my mailboxes, then send an email from the first one to bob@example.com saying the invoice is approved.
A typical agent loop for a support bot:
check_new_messages() → [messageUids]
get_message(uid) → body, attachments
list_thread_messages(threadId) → full conversation context
reply_email(uid, { text: ... }) → response lands in the same thread
Configuration
| Environment variable | Required | Description |
|---|---|---|
ENGAGELAB_EMAIL_SECRET_KEY |
Yes | Secret Key starting with sk_ |
ENGAGELAB_EMAIL_BASE_URL |
No | Override the API base URL. Inferred from the key region: sg → Singapore, tr → Türkiye |
Limits: up to 10 attachments / 10MB total per message (base64-encoded in the tool schema).
Troubleshooting
list_mailboxesreturns an empty list — your account has no mailbox yet. Create one in the EngageLab console (see Quickstart step 2). Agents cannot self-provision mailboxes yet.- 401 /
code 100101— the Secret Key is wrong or revoked. CheckENGAGELAB_EMAIL_SECRET_KEYand that the key's region prefix (sk_sg_/sk_tr_) matches the endpoint you use. - Could not determine API base URL — the key's region prefix isn't recognized. Set
ENGAGELAB_EMAIL_BASE_URLexplicitly. - Reply shows up as a new conversation — use
reply_email(notsend_email) so In-Reply-To headers and threading stay intact.
Roadmap
- [ ] Hosted MCP (OAuth) for zero-config remote use
- [ ] Programmatic mailbox creation via Secret Key
- [ ] Webhook push (no polling) for
check_new_messages - [ ] URL-based attachment references (no base64 in context)
Run from source
git clone https://github.com/Metaverse-Cloud/engagelab-email-mcp
cd engagelab-email-mcp
npm install && npm run build
ENGAGELAB_EMAIL_SECRET_KEY=sk_sg_xxx node dist/index.cjs
Related
- EngageLab Agent Email CLI — same mailboxes from the terminal,
--jsonoutput for agents - EngageLab Node SDK / Python SDK — send email from code
License
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.