I got an AI assistant (Claude) to partially manage my MLO tasks — and why a native MCP would change everything

[João Saleiro]

Hello,

I want to share something that has genuinely changed how I work with MyLifeOrganized — and, more importantly, make a case to the MLO team for a feature that would make it dramatically better. :)

What I can do now

Using Claude (Anthropic's AI assistant), I can create, update and review my MLO tasks by voice, from anywhere. While I'm driving, I can ask "what are my tasks for today?" or say "add a task to call the accountant tomorrow" — and it lands in MLO. For me, that hands-free, from-anywhere capture and review is a lifechanger.

How it works (high level)

I have MLO Cloud and the Google Calendar sync both enabled. From there:

• Claude connects to my Google Calendar to read the tasks, and to create or update them.
• Because MLO syncs both ways with Google Calendar, anything Claude writes to the calendar flows back into MLO automatically.
• To keep it cheap (no wasted AI usage), I had Claude generate a small Python script that mirrors the tasks into local Markdown files and refreshes them every ~2 minutes. So day-to-day reading costs nothing; the AI is only used when I actually want to review or change something.

I deliberately kept the setup simple and won't go into the technical weeds here — the point is that it's built entirely on top of the existing Calendar sync, with no official MLO integration.

The honest part: this is a workaround, and it shows

As happy as I am, this is far behind what would be possible if MLO exposed a proper integration. The current limitations:

• The Google Calendar sync is manual and desktop-only, so the whole chain depends on my desktop being on.
• The Calendar sync is slow.
• Each task carries only minimal information — title, date, recurrence. No notes, no subtasks, no contexts, no priorities, no folders. The AI is flying half-blind.

My ask to the MLO team

Please consider building an MCP server (or a proper public API) for MLO Cloud. MCP (Model Context Protocol) is quickly becoming the standard way for AI assistants like Claude to talk to apps directly. A native MLO MCP would remove every limitation above at once: real-time, cross-device, no Calendar detour, and full access to the rich structure MLO already has (subtasks, notes, contexts, priorities, due/start dates, recurrence, folders).

MLO's outline/hierarchy model is one of the best in the task-management world. Pairing that depth with an AI that can read and manipulate it directly would be a genuine differentiator — and, frankly, the workaround I've described is proof that the demand is already here.

Happy to share more with anyone who wants to try something similar in the meantime. :)

João Saleiro

[Alyona (MLO Support)]

Thanks for sharing this, João!  we really appreciate you taking the time to document your setup and provide such thoughtful feedback.

The good news is that we're already moving in this direction. As part of the MLO Web development, we're working towards a support for AI integrations. While it's still a work in progress, your use case is exactly the kind of scenario we want to enable more naturally in the future.

Thanks again for sharing your experience - it helps validate the direction we're taking.