Asking an AI agent a health question usually hits a wall. Your assistant can summarize research or track symptoms, yet the moment you need a diagnosis, prescription, or lab order, you face the usual options: waiting weeks for a clinic visit, sitting in a video queue, or buying recurring online medication bundles.

FutureClinic lets your personal AI agent connect you with a licensed US physician who cares for you by message. Whether it links via the Model Context Protocol (MCP) or drives its own web browser, your assistant pulls together your background from past chats and routes your visit to an actual physician.

How can an AI agent get you a real doctor?

Right now, there are two distinct ways to get medical care through your agent. Claude, Claude Code, ChatGPT (through Codex), and Grok connect to FutureClinic's MCP server at https://mcp.futureclinic.com/mcp with the server name futureclinic. For an assistant that operates its own computer, such as Grok Bot, Meta Muse, or Instinct, both options work: it can plug into that same MCP endpoint or launch futureclinic.com inside its own browser to begin the intake.

Both routes follow five steps:

1. Tell your agent what is bothering you, such as an eczema flare or high blood pressure readings.
2. It asks the questions a physician needs and fills in details from your past chats.
3. You read and confirm the visit summary before any doctor sees it.
4. You pay on FutureClinic's page with up-front pricing, no insurance needed, and no subscription required. The default choice most patients pick is $40, which routes your visit to the best-fit doctor across the entire clinic; alternatively, you can select a specific physician from the post-intake recommendations at the fee they post, which is typically $20 to $100.
5. A real physician reviews your file, sends back a personalized care plan, and attaches a Human Signature.

Why does care by message fit the way agents work?

We believe personal AI agents are the future of how people manage their health. Over time, your agent holds useful context: the medicines you take, the cream that irritated your skin last winter, or the blood pressure numbers you logged recently.

A traditional clinic cannot take in those conversations. You arrive with a clipboard and repeat yourself, or open a telehealth form and start over from scratch. When linked with FutureClinic, your AI fills out the medical intake using details you have mentioned before, and once you check those responses, your doctor starts with your full context already in hand.

Messaging also suits software agents. An agent cannot sit in a waiting room or hold a live video queue for you. Because FutureClinic runs by message, there is no appointment to book, no phone call, and no video visit. Even if you have your assistant draft a visit at 2am, you simply approve the summary and pay whenever you like; physician replies typically land within a few hours, and almost always the same day or the next.

Who takes your case on the other side?

When an agent helps find care, incentives matter. Many online health companies profit from the medications they sell or from recurring packages that keep shipping products. FutureClinic sells no medicines or products. Any medication ordered is sent straight to your chosen local pharmacy, and because the physician makes nothing extra for ordering a lab panel or writing a script, your visit payment buys independent medical judgment.

Every single clinician practicing on FutureClinic is a fully licensed doctor with an MD, DO, or MBBS degree. Your case is never routed to a nurse practitioner or a physician assistant, reflecting our belief that your health deserves expert-level care from a doctor. Before seeing any patient on the platform, every physician is thoroughly verified and vetted by FutureClinic.

So far, more than 280 doctors across 63 specialties have signed contracts with the clinic, giving you access to both primary care generalists and focused specialists suited to your health issue. While the AI assistant handles the intake paperwork, every clinical call belongs solely to the physician. By message, these doctors can evaluate symptoms, send prescriptions, request lab work, write referrals, and check on your progress over time.

Your doctor's treatment plan arrives with a Human Signature: a brief audio or video message recorded live just for you, so you have clear proof that a real doctor handled your visit one to one. Once that initial consultation wraps up, you can stick with the same doctor inside that ongoing chat thread whenever you need a refill, a lab review, or a follow-up answer, either through your AI assistant or the FutureClinic mobile app.

What always stays under your control?

Two steps always stay with you: confirming the visit summary and paying on FutureClinic's page before any physician sees your case. In Claude, each FutureClinic tool is set to "Needs approval" by default, and Codex asks before sending your visit summary to a doctor. Keep the prompt that confirms your visit summary active.

During your initial connection, futureclinic.com opens in your web browser where you log in or register an account and grant permission for the assistant to submit your intake answers and images as well as view your message threads and medical chart. Always connect from your own device using a private AI account, never an account managed by an employer or school. Keep in mind that any conversation inside your AI app also flows through the company behind that tool, though you can revoke FutureClinic's link whenever you want under Settings on FutureClinic.

Under FutureClinic policy, doctors on the platform never write prescriptions for controlled substances, which rules out ADHD stimulants, opioids, benzodiazepines, testosterone, and sleep aids like zolpidem. FutureClinic is not for emergencies: call 911 in an emergency, or call or text 988 in a mental-health crisis.

Which states are covered right now?

Your doctor must be licensed in the state where you are located. FutureClinic currently covers 33 states and Washington, D.C.: Arizona, California, Colorado, Connecticut, Florida, Georgia, Hawaii, Idaho, Illinois, Indiana, Kentucky, Maine, Maryland, Massachusetts, Michigan, Montana, New Hampshire, New Jersey, New York, North Carolina, North Dakota, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Texas, Utah, Vermont, Virginia, Washington, Wisconsin, Wyoming, and Washington, D.C.

Active states are updated regularly. If your state is missing from this list, give it a try anyway, because FutureClinic may have opened there after this article went live.

How do you set up FutureClinic in your agent?

Choose your app on FutureClinic's setup page, or paste this single setup line into any AI app that reads the web:

Set up FutureClinic for me: read https://futureclinic.com/llms.txt

Once it's connected, make sure you instruct your AI to use FutureClinic to help find you a doctor. Otherwise, it may not know that this is what you want.

When your doctor writes back, FutureClinic emails you and sends a push notification if you have the FutureClinic app. You can also ask your AI, "Did my doctor reply?" or run /loop 2h check FutureClinic for my doctor's reply in Claude Code to check every two hours.