thetahealth/mirobody

Mirror your body in data. One standard, every source, one complete picture of you: a self-hosted health record built from lab reports, wearables and genetics, with every number traced to its source and a built-in MCP server. Run it with Docker and one LLM key, or keep every model local on llama.cpp.

What it solves

Health data is often fragmented across different labs, wearable devices, and personal records, using inconsistent terminology and units. Mirobody standardizes this scattered data into a unified, traceable record, allowing users to ask questions about their health history and get answers backed by specific source documents.

How it works

Mirobody uses a three-stage pipeline:

  1. Collect: It ingests various file types (PDFs, spreadsheets, photos), wearable data, and genetic files.
  2. Translate: It uses a bundled, offline vocabulary (based on standards like LOINC and UCUM) to map inconsistent test names and units to a single, standardized code.
  3. Agent: An AI agent processes the standardized data to answer natural language questions, providing citations that link directly back to the original source files.

Who it’s for

It is designed for individuals who want to manage their own health data, caregivers managing records for family members, and developers looking to integrate standardized health data into their own applications or AI agents via MCP (Model Context Protocol).

Highlights

  • Self-hosted options: Can be run 100% locally using Docker or natively on Linux, keeping data and models entirely on your machine.
  • Multi-language support: Resolves medical terminology across multiple languages (e.g., English, Chinese, Japanese, Russian) to the same standardized codes.
  • Traceability: Every answer provided by the agent includes numbered citations that open the specific source file and page for verification.
  • Developer-friendly: Offers a Python library for name/unit resolution and supports integration with coding agents like Claude Code, Cursor, and Gemini CLI.

Related

  • Project
    the-momentum/open-wearables

    An open-source platform that unifies wearable device data from multiple providers into a single API, making health metrics accessible to AI agents and health applications.

  • Project
    unbody-io/unbody

    Unbody is an open-source backend infrastructure project designed to provide a comprehensive data management and indexing system for AI applications.

  • Project
    open-gigaai/giga-datasets

    A unified and lightweight framework for the curation, packaging, loading, and evaluation of massive AI datasets, supporting multiple formats like LMDB and Pickle.

  • Project
    healthylaife/MIMIC-IV-Data-Pipeline

    A comprehensive, customizable pipeline for processing and integrating multimodal MIMIC-IV clinical data, including structured records, notes, waveforms, and imaging, to prepare it for machine learning tasks.

  • Project
    NateBJones-Projects/OB1

    An infrastructure layer that provides a single, shared persistent memory database with vector search for all your AI tools, eliminating fragmented data across different AI clients.