Does completing a task in LMCP sync to all my devices?

Yes. Microsoft To Do syncs through Apple Reminders on your Mac, so when the AI marks a task complete locally it propagates to To Do on your phone, the web, and Windows like any normal sync.

Can the AI complete more than one task at a time?

Yes. You can ask the assistant to complete several finished tasks in one prompt, and it will call the tool for each matching item.

Do I need a Microsoft API key or login for this?

No. LMCP works through the local Reminders sync on your Mac, so there are no API keys, OAuth setup, or cloud credentials required. You just need Microsoft To Do syncing to Reminders.

Complete Microsoft To Do Tasks with AI on Mac

Mark Microsoft To Do tasks complete just by asking

The todo_complete_task tool lets your AI assistant mark a Microsoft To Do task as done, right from a chat. Because Microsoft To Do syncs through Apple Reminders on your Mac, LMCP flips the completion flag locally and the change rides the normal sync back to your To Do lists on every device. You never open the app, hunt for the task, or tick a box yourself.

A real prompt looks like this: "Mark the 'Send Q3 invoice to Acme' task as complete in my To Do." The AI finds the matching task and completes it. You can also say things like "Close out everything I finished on the groceries list" and let the assistant complete several at once.

LMCP is a free, native macOS MCP server that gives AI assistants access to your real Mac apps locally. Download LMCP and connect it once.

Which AI agents work?

LMCP exposes 150+ tools to any MCP-capable assistant. Desktop clients auto-configure over local stdio:

Claude Desktop
Cursor
VS Code (GitHub Copilot)
Windsurf
Zed

Web-based AIs connect through the LMCP Cloud Relay connector:

ChatGPT
Claude.ai (web)
Grok
Perplexity

The desktop clients pick up LMCP automatically once installed; the web AIs use the relay so they can reach the same local tools securely.

Automation

Completing a task is rarely the whole job, and todo_complete_task chains naturally with other LMCP tools. Ask the AI to list your open tasks, complete the ones you've finished, then create follow-ups: "Show my work list, mark the deployment task done, and add a task to write the release notes." Pair it with Mail to complete a task and reply to the related thread in one breath, or with Calendar so finishing a prep task also blocks time for the next step. Because the assistant orchestrates several tools in sequence, a multi-step cleanup becomes a single sentence.

Context

The AI works against your real task data, not a guess. It reads the actual task titles, lists, and due dates synced through Reminders on your Mac, so when you say "the invoice task" it matches the exact item you mean. There's no separate copy to keep in sync and no manual export — the assistant sees the same lists you see, and the completion you trigger shows up instantly in Microsoft To Do on your phone and the web.

Productivity

Closing tasks one by one is friction that adds up: switch apps, scan the list, find the item, click. With todo_complete_task you stay in the conversation where you're already working. After a meeting, tell the assistant which action items you knocked out and it clears them in seconds. End-of-day reviews go from a few minutes of clicking to a single prompt. For anyone living in their AI client, it keeps Microsoft To Do accurate without ever breaking flow.

Privacy & GDPR

Everything runs locally on your Mac. LMCP talks to Reminders and Microsoft To Do's local sync directly — no task data is sent to or stored on any LMCP server, and there are no API keys to manage. Your lists never leave your machine. This local-first design makes LMCP GDPR-compliant by architecture, because there's simply no cloud copy of your personal data to govern.