Skip to main content

​
What is Mirobody Health?

Mirobody Health is an open-source AI-native data engine that bridges rich health context and the latest AI capabilities, enabling deeply personalized and comprehensive health AI experiences. It connects to various devices, handles files (lab reports, medical documents), and provides autonomous agent capabilities with MCP/Tools/Skills support, along with out-of-the-box Auth/Frontend/MCP server functions – ideal for managing personal health data with privacy or building customized health AI applications.

Quick Start

Get up and running in under 5 minutes with Docker

Use Cases & Examples

See real examples with synthetic health data

Provider Integration

Learn how to integrate new health device providers

API Reference

Explore the complete API documentation

​
Key Features

Connect wearables, devices, and medical records (coming later). Bring Your Own Data Provider (BYOP) in hours using our integration framework. OAuth-secured connections to Garmin, Whoop, and more.
Supports multimodal documents: PDFs, Excel files, photos (PNG, JPEG, WebP, HEIC, HEIF), and audio (WAV, MP3, AIFF, AAC, OGG, FLAC). Enables agent access to context within files, from lab reports to handwritten notes.
Includes health-focused tools out of the box. Add custom tools using Python without complex JSON schemas or manual bindings. Connect to third-party MCPs. Directly load standard Claude Agent Skills (coming later).
Powerful agent harness replicates Claude Code’s autonomous capabilities locally. Leverages tools intelligently and utilizes relevant health context for deep research and insights.
Functions as an OAuth-enabled MCP Server connecting to Claude, Cursor, or custom clients. Also operates as a GPT app, enabling capabilities within ChatGPT.
All components are open-source and run on-premises with full privacy. Bring your own LLM providers via OpenRouter or first-party keys. One-line install with Docker.

​
Supported Providers

Garmin

Activity, sleep, heart rate, HRV, and more

Whoop

Recovery, sleep, cycles, workouts, body metrics

Custom Providers

Easy integration guide for adding your own

​
Tech Stack

Mirobody Health is built with modern, production-ready technologies:
  • Python 3.12+: Async/await patterns for high performance
  • FastAPI: Fast, modern web framework for building APIs
  • PostgreSQL: Robust relational database for data storage
  • Redis: High-performance caching and session management
  • Docker: Containerized deployment for consistency
  • OAuth 1.0 & 2.0: Industry-standard authentication protocols

​
Who is Mirobody Health For?

Individual Users

Anyone seeking a private, intelligent platform to manage and understand their health data

Developers

Developers building health AI applications for themselves or organizations

​
Real-World Examples

All examples below were generated using synthetic health data. Some examples also utilize additional Theta private tools.

​
Private Health Data Management & Chat

Organize scattered data from wearables, healthcare portals, PDFs, and paper reports spanning years into one coherent memory. Access your full health context via built-in chat or on external platforms like ChatGPT or Claude.

​
Deep Research for Personal Health

Connect personal health data with public knowledge from online searches, tools, or MCPs. Conduct deep health research with reports similar to Gemini/GPT’s deep research, but highly tailored to your situation.

​
All-in-One Health Data Engine

For developers, Mirobody Health serves as the data engine for AI health applications, providing essential infrastructure so you can focus on creating unique user experiences.

Wearable Manufacturers

Integrate your device and instantly offer AI chat features with device data

Research Applications

Add customized tools to deploy subject-facing research apps without extensive development

Health Apps

Build consumer health apps with a complete data and AI backend

Enterprise Solutions

Contact us for additional enterprise-level support and custom integrations

​
Next Steps

1

Get Started

Follow our Quick Start guide to deploy Mirobody Health locally in minutes using Docker.
2

Configure Providers

Set up OAuth credentials for the health providers you want to integrate. See the Configuration guide.
3

Explore the API

Review the API Reference to understand how to link users, pull data, and access health information.
4

Add Custom Providers

Follow the Provider Integration Guide to add support for additional health devices.

​
Contributing

We welcome contributions! When you add a new device, a new tool, or improve the framework in any way, we would like you to share that with the community.

GitHub Repository

View source code, report issues, and contribute

Contributing Guide

Learn how to contribute to the project

​
Community & Support

Discord Community

Join our community for discussions and support (Coming Soon)

GitHub Issues

Report bugs and request features

Email Support

Get technical support from our team

​
Hackathon Coming Soon

We will launch hackathon(s) with prizes in the coming weeks. Potential themes include:
  • Best health application use case
  • Tool design challenge
  • Best prompts to one-shot device integration
Stay tuned for announcement details!
Mirobody Health is built on the Mirobody framework. For core framework documentation, visit the Mirobody documentation.