# Nightly AI training journal from your smartwatch to Notion

> Every night we check your workouts, and if you trained, an AI writes a short diary entry and files it in your Notion training log.

- Workflow type: agent
- Services: Google Health, Notion
- Categories: Personal Productivity
- Published: 2026-07-12

## What it does

- Checks your Google Health workouts at 9pm every night, so you never have to log a session by hand.
- Reads how hard the workout actually was from your heart rate zones and calories, not just how long it lasted.
- Has an AI write a short two-paragraph diary entry that reads like something you'd write yourself, plus a stats table.
- Files the entry as a new page in your Notion training log so the whole year builds up in one searchable place.
- Skips rest days automatically, so your log stays clean.

## What you'll need

- A Google Health account connected to a Fitbit, Pixel Watch, or compatible tracker that records your workouts.
- A Notion workspace with a training log database where the daily entries will land.
- A rough idea of the properties you want in the log (date, activity, duration, zone minutes, perceived intensity).

## Prompt

Every evening at 9pm in my local timezone, check my Google Health account for exercise sessions logged today. Use the List Data Points operation on the exercise data type, with today's start and end times as the range.

If the response is empty, it was a rest day. Stop there and do nothing. Do not create anything in Notion.

If one or more exercise sessions came back, then for the same time range also call List Data Points on the time-in-heart-rate-zone data type and on the total-calories data type. Together these tell you what the workout was (activity type, duration, distance if applicable), how the effort was distributed across heart rate zones (peak, cardio, fat burn, warm up), and total calories burned.

Also call List Data Points on the exercise data type for the previous 7 to 14 days so you have a rough sense of my recent training. Do not write a page for those older days, just use them as context.

Now write a short training journal entry as if I'm writing it myself in a diary. Two short paragraphs at most, thoughtful and personal, no jargon. First paragraph: what the workout was and how hard it actually felt, based on where the time was spent across heart rate zones. If most of the time was in cardio or peak zones, call it a hard session. If it sat mostly in warm-up and fat-burn zones, call it easy or a recovery day. Reference the actual minutes in each zone. Second paragraph: how this session compares with my recent training. Note whether today was harder, easier, longer, or roughly typical.

Then create a new page in my Notion training log database using the Create a Page operation. The page title is the date and activity, for example "2026-07-11 Interval run". Include a small properties table with the key stats: Date, Activity, Duration, Total calories, Minutes in peak zone, Minutes in cardio zone, Minutes in fat-burn zone, Minutes in warm-up zone, and a Perceived intensity value (Easy, Moderate, or Hard) that you derive from the zone distribution. The two-paragraph diary goes in the page body.

Keep the tone like a real training diary, not a report. Do not use em dashes.

## How to customize

- Change the run time. 9pm suits most people, but early risers often prefer an early-morning entry for the previous day.
- Swap in a different Notion database, or add extra properties (mood, weather, notes) that the AI should fill in.
- Adjust the tone. Make it a coach's voice, a bare stats log, or keep it as a personal diary.

## FAQ

### Do I need a Fitbit or a Pixel Watch?

Any tracker that syncs to Google Health works. Fitbit and Pixel Watch are the most common, but third-party devices that write into Google Health will also produce workouts the journal can read.

### What happens on a rest day?

Nothing. If Google Health has no exercise sessions logged for the day, the workflow stops silently and no page is created in Notion. Your log only fills up on days you actually train.

### How does it decide whether a session was hard or easy?

It looks at how many minutes you spent in each heart rate zone. A session mostly in the peak and cardio zones gets described as hard. Mostly warm-up and fat-burn shows up as easy or recovery. That way the entry reflects effort, not just duration.

### Can it compare today's workout to my recent training?

Yes. The AI also pulls the last week or two of workouts so it can note whether today was harder, easier, longer, or roughly typical for you lately.

### Where does the entry actually go in Notion?

It creates a new page as a child of your training log database. The date and activity become the title, the key stats go into the database properties, and the two-paragraph diary lives in the page body.

Use this prompt in General Input: https://www.generalinput.com/prompts/nightly-ai-training-journal-from-your-smartwatch-to-notion