Epic MCP Server & Agent Skill

What is the Epic API, and what can an agent read from it?

FHIR R4 REST APIs covering USCDI data and more: about 60 resource types in the sandbox, including Patient, Observation, Condition, MedicationRequest and DocumentReference, plus Bulk FHIR export on Group and CDS Hooks.

• API: Epic on FHIR (FHIR R4, STU3, DSTU2) and open.epic interfaces (FHIR R4)
• Access: Issued by your administrator
• Base URL: https://fhir.epic.com/interconnect-fhir-oauth/api/FHIR/R4/ (sandbox); production base URLs are per organization, listed in the open.epic Endpoints directory
• Authentication: SMART on FHIR EHR launch (OAuth 2.0 authorization code); SMART standalone launch (OAuth 2.0 authorization code); SMART Backend Services (JWT)

Anyone can create an app on fhir.epic.com and test it in the free sandbox. Production use needs the app marked ready for production and each Epic customer organization to download its client ID.

OAuth 2.0 with SMART on FHIR. Sandbox endpoints: https://fhir.epic.com/interconnect-fhir-oauth/oauth2/authorize and /oauth2/token. Backend apps use the client_credentials grant with a signed JWT client assertion.

What can an agent read from Epic?

• Patients — Patient read and search, plus the $match and $summary operations listed in the sandbox CapabilityStatement.
• Lab results and vitals — Observation read and search; DiagnosticReport read and search.
• Problems and allergies — Condition and AllergyIntolerance read and search.
• Medications — MedicationRequest read and search.
• Clinical notes — DocumentReference read and search for clinical notes and documents.
• Population exports — FHIR Bulk Data Access: Group $export kickoff, status polling and file download.

What can an agent write to Epic?

• Clinical notes — DocumentReference create and update (Clinical Notes) files a note into the patient chart. Confirm with the user first.
• Observations — Observation create and update, as listed in the sandbox CapabilityStatement.
• Problems and allergies — Condition.Create and AllergyIntolerance.Create (Patient Chart).

Records are HIPAA protected health information (PHI). Access is limited to the scopes granted to the app and, in production, to organizations that have activated its client ID.

How do I connect Epic to an AI agent over MCP?

As of October 2026 we found no MCP server published by Epic itself.

The practical path is to host a small MCP server that wraps the Epic on FHIR (FHIR R4, STU3, DSTU2) and open.epic interfaces, or to give the agent the calls directly with the SKILL.md below.

On ibl.ai an MCP connection is three records, and a working integration always needs all three:

1. Register the server — POST …/mcp-servers/ with its URL, a transport (sse, websocket or streamable_http) and an auth_type of none, token or oauth2.
2. Create a connection that holds the credential, scoped to the whole organization (platform), to a single agent, or to each user (user).
3. Enable it on the agent — mcp-tool in tool_slugs and the server id in mcp_servers. Both lists are replaced, not merged, so read the settings first.

Epic supports per-user OAuth, so the server can use auth_type=oauth2 with auth_scope=user: each person connects their own Epic account in the chat the first time the agent needs it.

The field reference is in MCP server connections, and the /iblai-api-agent-mcp skill runs all three steps from a coding agent.

What does the Epic agent skill do?

The Epic skill is one SKILL.md file named iblai-connect-epic. It gives an agent the auth conventions, what to read, which writes need a person's confirmation, the ibl.ai MCP wiring, example prompts and guardrails.

Download the Epic SKILL.md

Save it as .claude/skills/iblai-connect-epic/SKILL.md, or in your agent's skills folder. The ibl.ai platform skills it calls come from one install:

This connector skill is a starting point maintained on ibl.ai, not part of the iblai/vibe repository.

It was checked against Epic's documentation on October 20, 2026. Test it against your own Epic setup before you rely on it.

What can I ask an Epic agent to do?

• “Pull this patient's active problem list and current medications from Epic.”
• “Show the last three hemoglobin A1c results for patient MRN 1234567.”
• “Draft a visit summary from today's encounter notes for my review.”
• “List allergies recorded for this patient and flag any added in the last year.”

How do I run an Epic agent on ibl.ai?

Agentic OS is where the agent runs on ibl.ai: it holds the Epic MCP connection, the model the agent uses, its instructions and its conversation history.

On ibl.ai you own all the code and the data, run it model-agnostic across any LLM, and pay with no per-seat pricing — so you can deploy anywhere, from your own cloud to a fully air-gapped network.

• AI agents for Healthcare — what ibl.ai deploys in this sector

Which ibl.ai agents work with Epic data?

• Clinical Support — Healthcare
• Care Coordination — Healthcare
• Documentation — Healthcare

Which ibl.ai platform skills does an Epic agent use?

• /iblai-api-agent-mcp — MCP connectors for ibl.ai agents, end to end
• /iblai-vibe-agent-mcp — Add the agent MCP tab (Model Context Protocol connector management with featured connectors, custom connectors, OAuth,…
• /iblai-api-agent-tool — Enable or disable an ibl.ai agent's tools via the platform API

What else should I know before connecting Epic to an AI agent?

Which other data sources can an AI agent connect to alongside Epic?

Which other Healthcare systems have a connector?

• athenahealth — Cloud EHR and practice management
• Availity — Payer-provider health information network
• HL7 FHIR — Healthcare data exchange standard
• MEDITECH Expanse — Electronic health record (EHR)
• Oracle Health (Cerner) — Electronic health record (EHR)
• Veradigm — Electronic health record (EHR)

Every connector, grouped by sector, is on the connectors hub.