# HasteHealth [Health: Active]

**Category:** 🧬 Biology & Bioinformatics  
**Repository:** https://github.com/HasteHealth/HasteHealth  
**GitHub Stars:** 86  
**npm Downloads (last month):** 6604  
**Views:** 5  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/hastehealth

## Description
The Open-Source FHIR Server for AI-Native Health Apps

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "hastehealth": {
    "command": "npx",
    "args": ["-y","haste-health"]
  }
}
```

## Documentation

## What HasteHealth MCP server does

HasteHealth provides an MCP interface to a health data platform built around FHIR R4. It is intended for AI-native health applications that need access to clinical records through an agent-compatible protocol. The same platform serves applications and analytics through a FHIR API and supports SQL-on-FHIR for working with FHIR data in relational query workflows.

Clinical information is stored as FHIR R4. The platform describes inbound support for Epic, Oracle Health, and HL7v2 data, giving teams a way to bring data from those healthcare systems into a FHIR-centered environment before making it available to applications, analytics, and AI agents.

## How it works

The project is implemented in Rust and can be operated as a self-hosted deployment. Its architecture includes a server, a background worker, PostgreSQL, and an administrative application. The worker handles search indexing and FHIR subscription processing after the server and database services are running.

Docker Compose provides the quickest local path. The top-level compose file starts PostgreSQL, the server, worker, and admin application together. After the containers become healthy, the admin application is available at the documented local hostname and includes a tenant and user created by the migration step.

For backend development, the repository provides separate commands for starting supporting services, launching the server, running the worker, and developing the frontend. A single `haste-health` binary also provides server, worker, and administrative CLI commands.

## Setup and configuration

A local deployment can be started by downloading the repository’s `docker-compose.yml` file and running:

```bash
curl -O https://raw.githubusercontent.com/HasteHealth/HasteHealth/main/docker-compose.yml
docker compose up
```

The project can also be installed through npm:

```bash
npm install -g haste-health
haste-health --help
```

For a one-off invocation, use `npx haste-health --help`. Published binaries are available for Linux on x64 and arm64, and for macOS on Apple Silicon. Linux builds are static; Windows users can run the binary through WSL.

Configuration is supported through `haste.toml` or environment variables. Server and worker settings are documented separately in the project. Administrative commands include database migration and tenant creation. The repository also publishes Docker images for the server and admin application.

## Tools and capabilities

- Exposes clinical data to AI agents through MCP.
- Stores clinical records using FHIR R4.
- Provides a FHIR API for applications and analytics.
- Supports OAuth2 and SMART on FHIR.
- Offers SQL-on-FHIR access patterns.
- Processes search indexing and FHIR subscriptions in a background worker.

HasteHealth is licensed under Apache-2.0 and can be self-hosted. A hosted option is also offered with a signed BAA.

_Full upstream README: https://allmcps.com/mcp/hastehealth/readme_