Skip to content

deak-ai/openehr-mcp-server

Repository files navigation

openEHR MCP Server

An MCP (Model Context Protocol) server designed to interface with openEHR REST APIs, specifically the EHRbase implementation. This server enables MCP clients like Claude Desktop to create compositions for openEHR templates and submit them to a server. For production grade EHR integrations you must use an AI Model where you can ensure data privacy.

See https://modelcontextprotocol.io/introduction for more information about MCP.

MCP Tools

  • openehr_template_list: List all available openEHR templates from the EHRbase server
  • openehr_template_get: Retrieve a specific openEHR template by its unique identifier
  • openehr_template_example_composition: Generate an example openEHR composition based on a specific template
  • openehr_ehr_create: Create a new EHR in the system
  • openehr_ehr_get: Retrieve an EHR by its ID
  • openehr_ehr_list: List all available EHRs in the system
  • openehr_ehr_get_by_subject: Get an EHR by subject ID and namespace
  • openehr_composition_create: Create a new openEHR composition in the Electronic Health Record
  • openehr_composition_get: Retrieve an existing openEHR composition by its unique identifier
  • openehr_composition_update: Update an existing openEHR composition in the Electronic Health Record
  • openehr_composition_delete: Delete an existing openEHR composition from the Electronic Health Record
  • openehr_query_adhoc: Execute an ad-hoc AQL query against the openEHR server

MCP Prompts

  • vital_signs_capture: Capture vital signs for a specific EHR ID

MCP Resource

Not yet implemented

Quick Start with Docker

The easiest way to get started is to use the pre-built Docker image available on Docker Hub.

1. Prerequisites

Ensure you have a running EHRbase server. For running one locally, see below.

2. Configure Claude Desktop

Edit your Claude Desktop configuration file (claude_desktop_config.json) and add an "openEHR" object inside the "mcpServers" object.

{
  "mcpServers": {
    "openEHR": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "--network=host",
        "-e",
        "EHRBASE_URL=http://localhost:8080/ehrbase/rest",
        "ctodeakai/openehr-mcp-server:latest"
      ]
    }
  }
}

You can point EHRBASE_URL to your own EHRbase server or use the provided docker-compose setup.

Setup

Prerequisities

For this MCP server to work, you currently require

Local EHRBase Setup

Prerequisites

  • A working Docker installation
  • Python 3 (this project was built with python 3.12, earlier versions might work)
  • A Python virtual environment (pip, conda or uv)
  • Install the required dependencies in your Python environment:
    pip install -r requirements.txt

Running the EHRbase Server

  1. Navigate to the docker-compose directory:

    cd docker-compose
  2. Start the EHRbase server in detached mode:

    docker compose up -d
  3. Check the logs to verify the server is running properly:

    docker compose logs -f
  4. The EHRbase server will be available at http://localhost:8080/ehrbase/

  5. The EHRBase API documentation should be here: http://localhost:8080/ehrbase/swagger-ui/index.html

Uploading the Vital Signs Template

After setting up the EHRbase server and your Python environment, you can upload the vital signs template:

python scripts/upload_template.py

You can also specify a custom template or EHRbase URL:

python scripts/upload_template.py --template path/to/template.opt --ehrbase-url http://custom-url:8080/ehrbase/rest

You should see output confirming the successful upload of the template to the EHRbase server.

Creating an EHR

After uploading the template, you need to create an Electronic Health Record (EHR) to store compositions:

python scripts/create_ehr.py

This will create an EHR with a randomly generated subject ID. You can also specify a custom subject ID:

python scripts/create_ehr.py --subject-id "patient_12345"

The script will output the EHR ID, which you'll need when creating compositions or using the MCP server.

Running the Integration Tests

To run the tests, you'll need to install the test dependencies first:

pip install -r requirements-test.txt

After installing the test dependencies and uploading the template, you can run the tests with:

python -m pytest tests/test_*.py -v

This will run all the tests in the tests directory.

Running the openEHR MCP server with Docker

Building the Docker Image

Build the Docker image from the project root:

docker build -t openehr-mcp-server .

Running the Docker Container

docker run -i --rm --network=host openehr-mcp-server

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published