loading…
Search for a command to run...
loading…
Enables AI applications to securely search and manage healthcare data from FHIR R4-compliant servers with built-in safety validation for AI-generated clinical o
Enables AI applications to securely search and manage healthcare data from FHIR R4-compliant servers with built-in safety validation for AI-generated clinical observations, preventing recording of physiologically impossible values.
Healthcare AI products handle Protected Health Information (PHI) under HIPAA in the US and special-category personal data under GDPR Article 9 in the EU. Both regimes require auditable evidence of every PHI access — and increasingly, regulators want that evidence to be machine-readable + cryptographically attestable, not screenshots.
FHIR R4/R5 is the de-facto interoperability standard now. Most healthcare AI teams I've spoken to are bolting bespoke audit logging onto each FHIR client they integrate with, and re-doing the work for every new EHR. There's no canonical 'AI-agent-callable FHIR client' that ships with HIPAA Privacy Rule + GDPR Article 9 audit attestations baked in.
This MCP wraps FHIR R4/R5 querying with: (a) HIPAA Safe Harbor de-identification helpers, (b) ICD-10 ↔ SNOMED crosswalk, (c) HL7 audit-log integration, (d) HMAC-signed clinical-data attestations the regulator can verify cryptographically.
A US-EU-dual-jurisdiction telehealth startup needed to give their AI agent safe access to patient observations across multiple FHIR-conformant EHRs (Epic, Cerner, NHS Spine). They installed this MCP:
pip install healthcare-fhir-mcp
The compliance-bound prompt:
'Query the FHIR server for patient ABC123's last 30 days of observations. Apply HIPAA Safe Harbor de-identification. Produce a clinical timeline. Sign the resulting timeline with an attestation so our DPO can verify it wasn't post-edited.'
Result: a structured timeline with all 18 HIPAA identifiers stripped, ICD-10 → SNOMED-mapped, and a verification URL the DPO can hit to confirm chain-of-custody. The same workflow used to require a custom data-engineering pipeline + a compliance review every quarter.
By MEOK AI Labs — Sovereign AI tools for everyone.
FHIR R4 MCP server for healthcare AI applications. Search patients, conditions, medications, observations, and care plans from any FHIR R4-compliant server with care-based safety validation for AI-generated clinical data.
MCPize MIT License MEOK AI Labs
| Tool | Description |
|---|---|
search_patients |
Search patients by name, date of birth, or identifier |
get_patient |
Get a full patient record by FHIR resource ID |
search_conditions |
Find diagnoses and conditions for a patient |
search_medications |
Find medication requests (prescriptions) for a patient |
search_observations |
Find lab results, vital signs, and observations |
create_observation |
Record a new observation (vital sign, lab result) |
get_care_plan |
Retrieve active care plans for a patient |
validate_resource |
Validate a FHIR resource against the R4 specification |
pip install mcp
git clone https://github.com/CSOAI-ORG/healthcare-fhir-mcp.git
cd healthcare-fhir-mcp
python server.py
{
"mcpServers": {
"healthcare-fhir": {
"command": "python",
"args": ["server.py"],
"cwd": "/path/to/healthcare-fhir-mcp"
}
}
}
| Plan | Price | Requests |
|---|---|---|
| Free | $0/mo | 100 requests/day |
| Pro | $15/mo | 10,000 requests/day |
| Enterprise | Contact us | Custom + HL7v2 bridge + SLA |
This is one of 255+ MCP servers by MEOK AI Labs. Browse all at meok.ai or GitHub.
| Plan | Price | Link |
|---|---|---|
| Healthcare FHIR MCP | £15/mo | Subscribe |
| Full Suite (9 MCPs) | £999/mo | Subscribe |
Built by MEOK AI Labs — sovereign AI infrastructure.
MEOK AI Labs | meok.ai | csoai.org | [email protected]
Добавь это в claude_desktop_config.json и перезапусти Claude Desktop.
{
"mcpServers": {
"healthcare-fhir-mcp-server": {
"command": "npx",
"args": []
}
}
}