Does list_emails work with Gmail and Outlook?

It works with any account configured in Apple Mail (Mail.app), including iCloud and Gmail. For Microsoft 365 / Outlook work mailboxes, use the m365_list_emails tool instead, which talks to your M365 account.

Does the AI read the full content of my emails?

No. list_emails returns headers only — subject, sender, date and unread status. To open the body of a specific message, the AI calls read_email separately, so it only reads full content when you ask it to.

Is my email data sent to the cloud?

No. LMCP runs natively on your Mac and reads Apple Mail locally. No headers or messages are stored on a server or copied to the cloud — everything stays on your machine.

List Apple Mail Emails with AI on Mac

The list_emails tool lets your AI assistant read the headers of messages in Apple Mail (Mail.app) on your Mac — subject, sender, date and unread status — for your local accounts like iCloud and Gmail. It returns headers only, so it's fast and gives the AI a clean overview of your inbox without dumping full message bodies. A typical prompt is simply: "Show me the unread emails in my inbox from today." Your AI calls list_emails, gets the headers back, and summarizes who emailed you and about what. (Using a Microsoft 365 / Outlook work account instead? Use m365_list_emails for those.)

LMCP is a free, native macOS MCP server with 150+ tools. It connects AI assistants to the real apps on your Mac — locally. No API keys, no cloud mailbox sync, nothing stored on a server. Download LMCP to get started.

Which AI agents work?

Desktop AI clients auto-configure through LMCP's local stdio connection, so they discover list_emails the moment you restart them:

Claude Desktop
Cursor
VS Code (GitHub Copilot)
Windsurf
Zed

Web-based AIs connect through the LMCP Cloud Relay connector, which securely bridges the assistant to your local Mac:

ChatGPT
Claude.ai (web)
Grok
Perplexity

Automation

Listing headers is the entry point to almost every email workflow. Because list_emails returns lightweight headers, the AI uses it to decide what matters, then chains to other tools: it can call read_email to open the body of a specific message, search_emails to narrow by sender or keyword, or reply_email and send_email to act. Pair it with Calendar and Reminders tools and the AI can turn an inbox scan into scheduled events and follow-up tasks — all in one prompt.

Context

Generic chatbots can only guess about your day. With list_emails, the AI works with your actual Apple Mail data — the real senders, real subjects and real timestamps sitting in your inbox right now. Ask "Did the accountant reply yet?" or "What came in while I was in that meeting?" and the answer is grounded in your live mailbox, not a stale or imagined version of it.

Productivity

Triage that normally means scrolling through dozens of messages becomes a single question. Instead of opening Mail.app and reading every subject line, you ask your AI to list and group what's unread, flag what looks urgent, and ignore newsletters. A morning inbox review that took ten minutes drops to a ten-second prompt — and because headers come back fast, the AI can scan large inboxes without slowing down.

Privacy & GDPR

Everything runs locally on your Mac. list_emails reads Apple Mail directly on your machine through macOS permissions — your email headers are never copied to a server, never sent to LMCP, and never stored in the cloud. The AI only sees what you ask for, in the moment you ask. This is privacy by architecture, which makes LMCP a strong fit for regulated work. Learn more in our guide on being GDPR-compliant by architecture.