Hand years of heart data to the AI of your choice.
The AI export is a plain-language, self-describing dataset — built so Claude, ChatGPT, or Gemini has real context about your Apple Health data. The app produces excellent data; the analysis is yours to ask for.
A dataset that describes itself
Every section names its exact HealthKit type, its units, and what the metric means, and the file opens with full provenance — so an assistant reads your history the way a researcher would, not as an unlabeled wall of numbers.
- ✓Provenance header — export version, generated time, source devices, locale, time zone, date range
- ✓Exact HealthKit identifiers and canonical units per metric
- ✓Objective statistics and monthly aggregates
- ✓ECG classification counts, passed through from Apple verbatim
- ✓Two depths: Summary (compact) or Complete (every sample + provenance)
# Heart Data Export for Claude
You are helping me analyze my long-term heart data. Below is my
cardiac data exported from Apple Health.
Export Version: 1.0
Generated: 2025-07-14T05:00:00Z
Source: Apple Health (on-device export)
Device(s): Apple Watch Series 10, Withings BPM
Locale: en_US
Time Zone: UTC (all timestamps are ISO-8601 UTC)
Date Range: 2025-01-15 through 2025-07-14
Readings: 120 · ECG records: 4 · Blood pressure readings: 3
Depth: Summary (statistics and monthly aggregates)
## Resting Heart Rate
HealthKit Identifier: HKQuantityTypeIdentifierRestingHeartRate
Unit: bpm (HealthKit unit: count/min)
Definition: Heart rate measured while at rest, as recorded by Apple Health.
Statistics: avg 60, median 60, min 56, max 64, sd 2.1, n=30, trend falling
Monthly averages:
2025-01: 62 (n=3)
2025-02: 62 (n=5)
...
## ECG
HealthKit Identifier: HKDataTypeIdentifierElectrocardiogram
Records: 4
Classification counts:
Sinus Rhythm: 3
Atrial Fibrillation: 1Analyze your Apple Health data with AI — on your terms
Large language models are good at reading long, structured records and surfacing patterns you might ask about — but only if the data arrives with context. That is what this export provides: a self-documenting file that states what each metric is, in what unit, over what time zone and range, from which devices. You bring it to Claude, ChatGPT, or Gemini and ask your own questions. Heart Data Exporter never inserts conclusions of its own — see our approach to interpretation.
Prefer a spreadsheet or a database instead? The same data is available as CSV, Excel, JSON, or SQLite. Need something for a clinician? Try a print-ready report.
Frequently asked questions
How do I get my Apple Health data into ChatGPT or Claude?
Choose the AI export in Heart Data Exporter, pick your assistant (Claude, ChatGPT, or Gemini) and a depth, then share the generated file — or copy its text and paste it into a conversation. The export is plain-language Markdown designed to give the assistant real context about your data.
What is the difference between Summary and Complete?
Summary is compact and token-efficient: per-metric statistics plus monthly averages, ideal for a conversation. Complete includes every individual sample as a table with full provenance — start and end time, value, source app, device, UUID, and metadata — for deep analysis or custom tooling.
Why is the AI export better than pasting a spreadsheet?
It is self-describing. Each section names the exact HealthKit identifier, the units, and a plain-language definition, and the file opens with provenance metadata — export version, date range, devices, time zone. An assistant reading it knows precisely what each number means instead of guessing from column headers.
Does the app analyze my data or tell the AI what to conclude?
No. The export contains data and definitions only — no interpretations, no health scores, and no suggested conclusions. What to ask, and how to read the answer, is entirely up to you and the assistant you choose.
Give your AI the whole picture.
Heart Data Exporter is a one-time $7.99 purchase. No subscription, no ads, no accounts — buy it once and it is yours.