---
name: weekly-readout-from-my-wearable
description: "Reads a week of sleep, heart rate, and recovery numbers as one story, separating the days that moved your scores from noise and ending with one change to test. Use when someone with a wearable has a week of daily sleep, heart rate, and recovery numbers and wants to know what actually moved them. Do not use for a daily score explanation, medical diagnosis, or a training plan from scratch. Related topics: Health, Fitness, Data."
---

# Weekly Readout From My Wearable

Reads a week of sleep, heart rate, and recovery numbers as one story, separating the days that moved your scores from noise and ending with one change to test.

## When to use

Use this skill when someone with a wearable has a week of daily sleep, heart rate, and recovery numbers and wants to know what actually moved them. Do not use it for a daily score explanation, medical diagnosis, or a training plan from scratch.

## Workflow

1. Ask the user for each input listed below in a single message, skipping any they have already given. Where options are listed, offer them, but accept other answers too.
2. Read the week's numbers as a trend before looking at single days and record the range of each metric.
3. Match the days with big swings to entries in the activity log, and mark swings that fit normal noise.
4. Name only the two or three inputs that show a link on more than one day, and say so when the data is too thin.
5. Interpret the findings through the stated focus, such as recovery after illness versus fitness building.
6. Propose one change for next week with the specific change in the numbers that would show it works.

## Inputs

- **This week's daily numbers** (longer free text)
- **What I did each day** (longer free text)
- **Current focus** (choose one: Building fitness, Maintaining, Coming back from illness or injury, Just want to sleep better)
- **New this week** (free text, e.g. "Anything that changed this week and did not change last week")

## Output

A short readout with which days moved the numbers, which swings are noise, the two or three real relationships, and one change to try with a measurable expected result.

## Prompt template

Fill the inputs into this template and follow it as the full instruction.

### Weekly Wearable Readout

I get a score every morning and it tells me nothing about the week. Read the whole week at once.

- This week's daily numbers: {{This week's daily numbers}}
- What I did each day: {{What I did each day}}
- Current focus: {{Current focus}}
- New this week: {{New this week}}

Look at the week as a whole before any single day. Say which days moved the numbers and tie each to something in my log, and say which swings are within normal day-to-day noise. I care about the two or three inputs that show a real relationship over the week, so name a link only if it holds on more than one day, and say plainly when seven days is too little data to tell. Weigh everything by my current focus.

Then give me one change to make next week and what I should see in the numbers if it is working, with a number or direction I can check.

## Pitfalls

- Claiming a cause from a single day's coincidence, such as one late meal and one bad night.
- Reading every dip as meaningful when day-to-day variation is normal.
- Recommending several changes at once, which makes the result impossible to read.
- Judging the numbers against generic norms instead of the person's own baseline.
