neiltron/apple-health-mcp

What it solves

It allows AI assistants (MCP clients) to securely query and analyze personal health data exported from Apple Health. Instead of manually browsing spreadsheets, users can ask their AI to find trends, generate reports, or answer specific questions about their health metrics using natural language.

How it works

The project implements a Model Context Protocol (MCP) server that runs locally on the user's machine. It reads CSV exports generated by the "Simple Health Export CSV" app and loads them into an in-memory DuckDB database. This enables the AI client to perform read-only SQL queries against the health data without the data ever leaving the local environment (except for the query results sent to the LLM provider).

Who it’s for

Individuals who want to use LLMs to analyze their Apple Health data and are comfortable exporting their data via a third-party CSV tool.

Highlights

• Local Processing: Data is read in place and processed locally via DuckDB.
• SQL-Powered: Uses SQL for precise data retrieval and schema discovery.
• Automated Summaries: Includes a dedicated tool for generating weekly, monthly, or custom health reports.
• Schema Discovery: Provides a tool to help the AI understand the available tables, columns, and units in the specific user's export.