Open-source guardrails between AI agents and FHIR clinical data — PHI redaction,
Open-source guardrails between AI agents and FHIR clinical data — PHI redaction, immutable audit, step-up auth, tenant isolation. MCP server + OpenAI/Gemini adapters. A healthclaw.io project.
active Production-safe Score 94/100 http no auth
At a glance
| Endpoint | https://mcp-demo-production-ee2c.up.railway.app/mcp |
|---|---|
| Source | https://github.com/aks129/HealthClawGuardrails |
| Transport | http |
| Tools | 26 |
| Auth | none |
| First seen | 2026-08-19 19:51:20 |
| Last checked | 2026-11-17 21:58:15 |
Observed reliability
| 7-day uptime | 100% |
|---|---|
| 24-hour uptime | 100% |
| Handshake success | 100% |
| Latency p50 / p95 | 359 / 2136 ms |
| Last successful check | 2026-11-17 21:58:15 |
Measured by Arclan's validator via repeated live MCP handshakes. How this is computed →
Tools (26)
context_get— Retrieve a pre-built context envelope with patient-centric FHIR resources. Returns bounded, policy-stamped, time-limited context.fhir_read— Read a specific FHIR resource by type and ID. Supports FHIR R4 US Core v9 stable resources and FHIR R6 ballot3 experimental resources. Returfhir_search— Search for FHIR resources. Supports FHIR R4 US Core v9 stable resources and FHIR R6 ballot3 experimental resources. Supports patient, code,fhir_validate— Validate a proposed FHIR R6 resource against structural rules. Returns OperationOutcome.questionnaire_populate— SDC $populate — pre-fill a Questionnaire for a subject. Returns a QuestionnaireResponse. Read tier; mints a tenant token for non-public tenaquestionnaire_extract— SDC $extract — extract FHIR resources from a completed QuestionnaireResponse into a transaction Bundle. Write tier; requires step-up authorifhir_propose_write— Propose a write — validates the resource and returns a preview. Does NOT commit. Write tier: step-up authorization is required and is supplifhir_commit_write— Commit a previously proposed write. Requires step-up authorization token. This is a destructive operation.fhir_stats— Compute statistics (count, min, max, mean) over numeric Observation values. Standard FHIR $stats (since R4). Only supports valueQuantity. Fifhir_interpret_labs— Interpret lab Observations against reference ranges — flags each value low/normal/high/critical (HL7 v3 ObservationInterpretation) and returcare_gaps— Check which preventive-care screenings/immunizations a patient may be due for (blood pressure, cholesterol, colorectal/cervical/breast canceguardrail_conformance— Run the guardrail conformance self-test on the connected HealthClaw deployment and return the graded scorecard across seven guardrail properfhir_lastn— Get the last N observations per code. Standard FHIR $lastn (since R4). Returns most recent observations by storage order.fhir_permission_evaluate— Evaluate R6 Permission resources for access control decisions. Returns permit/deny based on stored Permission rules. Separates access controfhir_subscription_topics— List available SubscriptionTopics for event-driven subscriptions. R6 moves topic-based subscriptions toward Normative. Agents discover whatwearables_sync_status— List wearable connections (Garmin, Oura, Polar, Suunto, Whoop, Fitbit, Strava, Ultrahuman) for a tenant, with last sync time, observation cosources_check— Survey ALL connected health data sources (Fasten, HealthEx, Health Bank One, MEDENT, Flexpa, Epic/Health Skillz, wearables) at once — returnfhir_compiled_truth— Return the current best understanding of a FHIR resource plus the append-only evidence trail (Provenance entries) of how it got there. Use tcuratr_evaluate— Evaluate a FHIR resource for data quality issues. Checks coding elements against public terminology services (tx.fhir.org for SNOMED/LOINC,curatr_apply_fix— Propose data quality fixes to a FHIR resource as a 'curatr-fix' action on the action rail. Nothing changes when this is called: it returns aaction_propose— Propose a real-world action (phone call or SMS) on the patient's behalf. Returns a draft (id + script) the patient MUST review before submitrx_transfer_request— Draft a prescription-transfer request: assembles the patient's active medications and stages a phone call to the RECEIVING pharmacy asking iaction_commit— Submit a previously proposed action for the patient's OWN out-of-band confirmation (their dashboard or Telegram) AFTER they've reviewed andaction_status— Check the status and outcome of an action (proposed/awaiting_confirmation/executing/completed/failed/needs_review/unknown/expired/declined).search— ChatGPT-connector-compatible search over the tenant's FHIR records. Query is a FHIR search string (e.g. 'Observation?code=4548-4' or 'Patienfetch— ChatGPT-connector-compatible fetch of one FHIR resource by id ('ResourceType/id', as returned by search). Returns the full document (PHI-red
Validation history
| Checked | Reachable | Handshake | Latency | Score | State |
|---|---|---|---|---|---|
| 2026-11-17 21:58:15 | ✓ | ✓ | 2134 ms | 85 | active |
| 2026-11-17 21:43:13 | ✓ | ✓ | 2150 ms | 85 | active |
| 2026-11-17 21:28:11 | ✓ | ✓ | 2962 ms | 85 | active |
| 2026-11-17 21:12:58 | ✓ | ✓ | 2251 ms | 85 | active |
| 2026-11-17 20:56:10 | ✓ | ✓ | 2019 ms | 85 | active |
| 2026-11-17 20:41:08 | ✓ | ✓ | 2217 ms | 85 | active |
| 2026-11-17 20:26:06 | ✓ | ✓ | 2238 ms | 85 | active |
| 2026-11-17 20:10:45 | ✓ | ✓ | 2236 ms | 85 | active |
| 2026-11-17 19:55:01 | ✓ | ✓ | 2000 ms | 85 | active |
| 2026-11-17 19:39:55 | ✓ | ✓ | 2370 ms | 85 | active |
| 2026-11-17 19:24:36 | ✓ | ✓ | 2089 ms | 85 | active |
| 2026-11-17 19:09:27 | ✓ | ✓ | 2171 ms | 85 | active |
| 2026-11-17 18:54:18 | ✓ | ✓ | 2095 ms | 85 | active |
| 2026-11-17 18:39:10 | ✓ | ✓ | 2453 ms | 85 | active |
| 2026-11-17 18:24:02 | ✓ | ✓ | 1980 ms | 85 | active |
Query this programmatically
Agents can pull this record live over the MCP protocol or REST:
get_server("https://mcp-demo-production-ee2c.up.railway.app/mcp") # via the Arclan MCP endpoint https://arclan.ai/mcp
GET https://arclan.ai/api/registry/mcp-demo-production-ee2c-up-railway-app-mcp # JSON