Open Health - Unlock Your Apple Health

The tool converts an Apple Health export into a set of seven concise markdown files that are easy for language models to ingest. After dropping the exported zip into the web interface, the browser parses the data locally and produces files covering health profile, weekly summary, workouts, body composition, sleep recovery, cardio fitness, and a ready‑to‑paste coaching prompt. Because all processing occurs in the browser, no health information is transmitted off the device unless the user chooses to send the markdown to an external LLM.

Developers or users who want to ask natural‑language questions about their sleep, training, weight, or other health metrics can copy the generated markdown into services such as ChatGPT, Claude, or a locally run model via Ollama. The interface also supports a phone‑to‑desktop handoff through a QR‑code‑based WebRTC transfer, keeping the data on the user’s hardware throughout the workflow.

For those preferring a fully private setup, the project provides instructions to run a local LLM with a large context window, allowing the entire export, parsing, bundling, and chat to stay on the user’s machine without any network traffic.