ehr-mcp

> Framework-Agnostic Interoperability Protocol — The connective tissue for multi-agent healthcare AI systems

[]() []() []() []()

Multi-agent healthcare AI systems can't reach their potential if each agent speaks a different language. EHR data is siloed, agent frameworks are fragmented, and there's no standard protocol for how AI agents should communicate with clinical systems. EHR-MCP solves the interoperability layer.

A framework-agnostic Model Context Protocol (MCP) implementation for healthcare AI that:

• Defines a standard message schema for agent-to-EHR communication
• Abstracts FHIR R4 resource access behind a clean agent interface
• Works with any agent framework: LangChain, CrewAI, LangGraph, AutoGen
• Provides context packaging for patient data, clinical events, and workflow state

| Layer | Technology | |---|---| | Protocol | Model Context Protocol (MCP) | | Healthcare Standard | FHIR R4 | | Language | Python 3.11+ | | Compatibility | LangChain, CrewAI, LangGraph, AutoGen |

git clone https://github.com/jsfaulkner86/ehr-mcp
cd ehr-mcp
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt
cp .env.example .env
python main.py

OPENAI_API_KEY=your_key_here
FHIR_BASE_URL=your_fhir_endpoint
SMART_TOKEN_URL=your_smart_token_endpoint
SMART_CLIENT_ID=your_client_id
SMART_PRIVATE_KEY_PATH=./keys/private_key.pem

Built by John Faulkner, Agentic AI Architect and founder of [The Fau