What is an AI digital twin?

A structured, reusable profile of you — your writing voice, key relationships, weekly rhythm, recurring themes and tone — synthesized by an AI from your own data so it can act, write and answer the way you would.

Does building a digital twin upload my data to the cloud?

No. LMCP reads your email, messages, calendar and contacts locally on your Mac and writes the twin to a file on your machine. Nothing is uploaded and there is no account — unlike cloud 'clone yourself' services that ingest your data.

What data does the digital twin use?

Mainly your sent email and your own messages (your real writing voice), plus Contacts, Calendar, Reminders and Notes to capture relationships, routines and recurring themes.

Build Your AI Digital Twin on Mac — Privately, From Your Own Data

An AI digital twin is a rich, reusable profile of you — how you write, who you talk to, how your week runs, and what you care about. Most "clone yourself" services build one by uploading your data to their cloud. LMCP builds yours entirely on your Mac. Your AI assistant reads across the apps you already use — Mail, Messages, Calendar, Contacts and Notes — and synthesizes a "Digital Twin" document that stays on your machine. Try a prompt like: "Build my digital twin from my Mac."

It runs through LMCP, a free, native macOS MCP server that lets AI assistants use your real apps — with no cloud upload and no API keys.

What goes into your digital twin?

The twin is grounded in your actual data, not a generic bio. Your AI samples across sources to find patterns:

Your voice — read from your sent email and your own messages: greetings, sign-offs, length, formality, emoji, the languages you switch between. This is the single most valuable signal, and it only exists in what you've actually written.
Your relationships — from Contacts and who you talk to most: your inner circle, work vs. personal, how you address different people.
Your rhythm — from Calendar and Reminders: recurring meetings, working hours, focus blocks, how you manage commitments.
Your themes — the projects, decisions and topics that recur across mail, notes and reminders.
Your tone & values — how you make asks, say no, give praise, escalate.

What can you do with your digital twin?

Once the twin exists, it makes every other AI task sound like you. Ask your assistant to "draft this reply in my voice," "answer this the way I would," "write my morning brief the way I'd write it," or "brief my new assistant on who I am and how I work." Because the twin was built from your real data, the output matches how you actually communicate — not a bland AI default.

Is my data safe? Where does the digital twin live?

Everything stays on your Mac. LMCP reads your email, messages, calendar and contacts locally, and writes the Digital Twin profile to a file (or a note) on your machine. Nothing is uploaded — there is no cloud copy of your data and no account to create. That's the core difference from cloud "AI clone" tools: with LMCP, the data that describes you never leaves the device. This makes it private by architecture. The skill also summarizes sensitive content into patterns rather than copying private messages verbatim, and you can update or delete the twin any time.

Which AI assistants can build a digital twin?

The digital-twin skill works across every major MCP-capable assistant:

Desktop clients — Claude Desktop, Cursor, VS Code (GitHub Copilot), Windsurf and Zed auto-configure via local stdio.
Web AIs — ChatGPT, Claude.ai (web), Grok and Perplexity connect through the LMCP Cloud Relay connector, which bridges the cloud assistant to your local Mac. Even then, your personal data is read on your machine — only the request and the assembled profile pass through.

Ready to meet your AI twin? Download LMCP and ask your assistant to build your digital twin.