One number for today, a week of numbers ahead, and an AI that answers questions about your own data in plain words. The risk score is built from what you log and what the app detects, weighted day by day as each trigger's effect fades.
The score starts from your triggers. Every trigger you log, and every one the app detects on its own from your wearable, your phone or the weather, gets a severity you chose during setup: reliably triggers, contributes, might play a role, or no influence. Recent triggers weigh more than older ones and the score decays over 7 days, so yesterday's late night still counts today, less so on Thursday.
Prodromes, your personal early warning signs, add to it. So does your menstrual cycle, with its own weights for the days before, during and after your predicted period. All of it is yours to adjust in Settings under Risk Model, and the weekly AI calibration nudges those values as the app learns from your data.
Because weather comes with a forecast and your period is predicted, the app can put the days ahead on the same scale as today. A pressure drop due on Thursday and a period due on Saturday both show up before they land, so you can move the meeting, book the lie-in or pack the medication.
The outlook is part of Premium. The free app shows today's score.
Ask MigraineMe is a chat that reads your health data and answers questions about it. Ask whether weather was a factor in your last migraine and you get the date, the pressure change that day, and whether that kind of drop has coincided with your migraines before. It reads your data; it does not log for you.
It is a place for questions, not diagnoses. Everything it says comes from your own history, so the more you log, the better its answers.
No. The score works from what you log and what your phone can detect. A wearable or Health Connect adds sleep stages, HRV, recovery and stress automatically, which gives the score more to work with.
The AI needs about five logged attacks before it has anything to say. Findings sharpen with more days of data.
Yes. Each trigger's severity lives in Risk Model and Manage Items, and you can edit it at any time. The weekly calibration only proposes changes; you accept or reject each one.