lmcp_welcome: See What Your AI Can Read on Your Mac

lmcp_welcome is the fastest way to confirm that your AI assistant can actually see your Mac. When you call it, LMCP returns a live snapshot of what it can read right now: how many events are on today's calendar, which reminders are due, the total number of contacts on the device, and how many unread emails are sitting in your inbox. It's a read-only health check and a friendly first hello rolled into one.

A concrete example of what you type into your AI: "Run lmcp_welcome and tell me what's on my plate today." The assistant calls the tool and replies with something like "You have 3 calendar events, 2 reminders due, 1,240 contacts, and 7 unread emails." No setup screen, no account, no copy-pasting tokens — just a real read of your machine.

LMCP is a free, native macOS MCP server with 150+ tools. It lets AI assistants use your real Mac apps locally. There are no API keys to manage, nothing is sent to a cloud service, and nothing is stored on a remote server. Download LMCP to get started.

Which AI agents work?

lmcp_welcome works across every major MCP-capable assistant:

• Desktop clients — Claude Desktop, Cursor, VS Code (GitHub Copilot), Windsurf and Zed auto-configure through a local stdio connection. LMCP wires itself into their MCP config so the tool appears automatically.
• Web AIs — ChatGPT, Claude.ai (web), Grok and Perplexity connect through the LMCP Cloud Relay connector, which securely bridges the web assistant to the server running on your Mac.

Either way, the data never leaves your machine — the relay only forwards the request to your local server and the answer back.

Automation

lmcp_welcome is the natural opening move in a chain. Because it surfaces counts in a single call, your AI can use it to decide what to do next: see 7 unread emails, then call list_emails to triage them; spot 2 reminders due, then call list_reminders for details; notice 3 events today, then call list_calendar_events to read the agenda. One snapshot becomes a branching workflow — "check my day, then draft replies to anything urgent" — without you naming each tool.

Context

The AI is working with your real Mac data, not a guess or a stale export. The calendar count comes from Calendar.app, reminders from Reminders, the contact total from your address book, and unread mail from Mail.app. When the assistant tells you what's on your plate, it's reading the same database your native apps read, in real time.

Productivity

Instead of opening four apps every morning to take stock, you ask one question and get a unified briefing in seconds. lmcp_welcome turns the "what does my day look like?" ritual into a single prompt, and because it chains into deeper tools, the same conversation can move from overview to action — reading, drafting, and scheduling — without ever leaving your AI window.

Privacy & GDPR

Everything runs locally on your Mac. lmcp_welcome reads your calendar, reminders, contacts and mail directly through native macOS APIs and returns the result to your AI client. There is no cloud copy of your data, no server-side storage, and no third-party account. This local-first design makes LMCP GDPR-compliant by architecture — your personal data simply never leaves the device.