apple-health-mcp
Loads Apple Health export data into a local SQLite database and exposes tools to query health metrics and workout records via natural language.
README
apple-health-mcp
MCP server that loads your Apple Health export into an in-memory SQLite database and exposes it to LLMs via the Model Context Protocol.
Quick start
uvx apple-health-mcp --input ~/Downloads/export.zip
Or install it permanently:
uv tool install apple-health-mcp
apple-health-mcp --input ~/Downloads/export.zip
Exporting your data from Apple Health
- Open the Health app on your iPhone
- Tap your profile picture (top-right)
- Scroll down and tap Export All Health Data
- Confirm — this may take a few minutes
- Save or AirDrop the resulting
export.zipto your Mac
Claude Desktop integration
Add this to your claude_desktop_config.json
(macOS: ~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"apple-health": {
"command": "uvx",
"args": [
"apple-health-mcp",
"--input",
"/absolute/path/to/export.zip"
]
}
}
}
Restart Claude Desktop. The tools will appear in the tools menu.
MCP tools
list_metrics()
Returns every distinct record type and its count.
list_workout_types()
Returns every distinct workout activity type and its count.
summary(metric, period)
Aggregates a metric by day, week, or month. Returns count, avg, min, max, and sum.
summary("StepCount", "week")
summary("HeartRate", "month")
query(sql)
Run arbitrary read-only SQL against the database. Only SELECT statements are allowed.
Schema:
records(type, source_name, unit, value, start_date, end_date)
workouts(activity_type, source_name, duration, duration_unit,
total_energy_kcal, total_distance, distance_unit, start_date, end_date)
Examples:
-- Daily step totals for the last 30 days
SELECT substr(start_date, 1, 10) AS day, sum(CAST(value AS REAL)) AS steps
FROM records WHERE type = 'StepCount'
GROUP BY day ORDER BY day DESC LIMIT 30;
-- Longest runs
SELECT start_date, duration, total_distance, distance_unit
FROM workouts WHERE activity_type = 'Running'
ORDER BY total_distance DESC LIMIT 10;
Environment variable
Instead of --input, you can set:
export HEALTH_EXPORT_PATH=~/Downloads/export.zip
Development
git clone https://github.com/smarzola/apple-health-mcp
cd apple-health-mcp
uv sync # install deps + dev tools
uv run pytest # run tests
uv run apple-health-mcp --input ~/Downloads/export.zip
Lint, format & typecheck
uv run ruff check src/ tests/ # lint
uv run ruff format --check src/ tests/ # format check
uv run mypy src/ # typecheck
Auto-fix:
uv run ruff check --fix src/ tests/ # auto-fix lint
uv run ruff format src/ tests/ # auto-format
Pre-commit hooks
uv run pre-commit install # one-time setup
# Now ruff + mypy run automatically on every commit
Cutting a release
gh release create v0.2.0 --generate-notes
# Triggers GitHub Actions → builds → publishes to PyPI → smoke tests
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。