Use a health agent to prep visits and draft follow-ups

A personal OpenClaw setup read FHIR health data to brief appointments, draft messages and appeals, and watch for trends.

Done with OpenClawยท Advanced

What they did
The writer had Claude Code build a guardrail layer between OpenClaw and a private HAPI FHIR server. That server combined Epic records from two health systems, pharmacy data, and Flexpa insurance claims. The proxy hid protected health information, kept logs of access, and stopped any clinical or money-related step until a person approved it. After that, the agent helped prepare for appointments, write messages and appeals, fill out forms, and monitor changes over time.
How it went
In one case, a colonoscopy was coded as diagnostic and triggered a $1,000 deductible. The agent walked through the billing, drafted a dispute, and considered whether to escalate, but the source does not say how that dispute ended.
Tips
Keep the agent on a separate machine with its own account, and avoid adding third-party skills. Make privacy filtering happen in the data pipeline rather than relying on the model alone.

Try it yourself with OpenClaw

Connect to my health records from [FHIR server, export file, or health app] in read-only mode and help me prepare for my [appointment type] on [date]. Review recent results, key trends in [metric], current medications, billing issues, and any open questions. Then draft a short list of questions for my clinician, plus a follow-up message or appeal draft if needed. Pause and show me the exact text before you send anything, submit any form, delete anything, or make any payment. End with a one-page summary I can print.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on evestel.substack.com โ†—

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.