How to Use the Function Health Connector in Muse

1. The short answer

As of October 25, 2026, Function Health is a live connector in Muse. Connect your Function account and Muse can read your lab results and clinician reviewed health summaries, then use them as context for your health goals, your plans, and your follow up. That is the whole idea: your AI stops giving generic health advice and starts reasoning from your actual biology.

Function Health was live at launch on September 8, 2026, announced by Function that day through PR Newswire, and named by Alexandr Wang among the launch connectors. It was still counted among the 54 connectors in the web app in the October 15 in app check. There has been no October announcement changing how it works. This is the connector for the moment you want Muse to know what your bloodwork actually says before it suggests what to eat, how to train, or what to ask your doctor.

Here is the honest version. This connector makes Muse dramatically more useful for health, and dramatically more sensitive, because lab results are among the most private data you own. The value is real: a plan built on your actual numbers beats a plan built on averages. The risk is real too: health data in a chat deserves more care than any other connector on this page. The rest of this guide covers how to connect it, the setup that keeps your health data safe, the uses that are actually worth your time, and the limits to know before you lean on it.

Muse is invite only right now. Need an invite code? Click here to grab one free from the community pool. Waiting on a launch in your country? Join the waitlist and we will send you a fresh code on launch day. Muse is available in the United States, Canada, and Japan.

2. What you need

• Muse access. You need an account with access first, in the United States, Canada, or Japan, where Muse has launched.
• A Function Health membership. The connector reads your Function account, so you need to be a Function member with lab results already in your account. A connector with no results behind it has nothing to show you.
• Your Function sign in. You connect through the Function authorization screen Muse shows you. Keep your password to yourself and sign in only on that screen.
• Approvals set to ask me. Before you connect anything, set approvals so nothing is shared or acted on without your tap. This is the single most important step on this page, because this connector touches your health data.
• A health goal in one sentence. Understand my cholesterol numbers. Build a training plan around my actual recovery markers. Track whether my iron is improving. One clear goal beats a vague request to look at my health.

3. How to connect it, step by step

The connection itself takes a few minutes. The setup around it matters more, so run the steps in order and do not skip the approvals step at the end.

1. Check your Function results first. Open your Function account and confirm your latest lab results are there. The connector is only useful once there are real results to read.
2. Open Muse and go to Settings. The connectors list lives there. What is printed there is what your account has, because rollout is staged and two accounts can see different lists on the same day.
3. Find Function Health in the connectors list and open it. If Function Health is not listed, your account has not received it yet. There is nothing to install and nothing you broke. Check again in a few days.
4. Sign in to Function. Use the Function account that holds your lab results. Remember this is the account the connector will use, so sign in with the right one the first time.
5. Grant the access Muse asks for. Muse needs permission to read your lab results and clinician reviewed summaries. Read the permission screen before you accept it. If a scope surprises you, stop and check what it covers before continuing.
6. Set approvals to ask me. In Muse settings, set approvals so that sharing or acting on health data always waits for you. This is what stops Muse from ever passing your results somewhere you have not approved.
7. Test with a read only request. Ask Muse how one lab result compares to your optimal range, and read the answer against your Function summary. If it matches, the connection is good. Do not ask it to build a plan on the first day.

Copy this prompt

You are my health context assistant. Use my Function Health connector only.
Rules: you are not a doctor and you never diagnose, prescribe, or tell me to start or stop any treatment. You read my results, you explain them in plain words, you help me prepare questions for my doctor. You never share my health data with anyone without my explicit approval.
My goal: [one sentence, e.g. understand my cholesterol numbers]
First, do this in order.
1. Tell me which of my recent results sit inside their optimal range and which sit outside it, in plain words.
2. For each result outside its range, explain what it means and what usually moves it, without telling me what to do.
3. End with the 3 questions I should ask my doctor, in order.
Do not diagnose anything. Do not prescribe anything.

4. What it actually does

The connector earns its place with four uses. Everything else is a variation on one of them.

Understand your lab results in plain words

Ask Muse how a lab result compares to your optimal range, and you get an explanation you can actually read, instead of a row of numbers and a reference range you do not understand. Function says Muse draws on your lab results and clinician reviewed summaries to inform the goals and tasks you bring to it. The value is not the number itself. The value is knowing which numbers are fine, which ones deserve a question, and which ones are trending the wrong way across your last two rounds of labs. A result that moved from inside its range to outside it is a result worth asking about, and a chat check catches that movement without you building a spreadsheet.

Build a health plan around your actual numbers

Give Muse a goal and your real results, and it can build or adjust a plan toward that goal using your own Function insights as context. That is the use Function itself names: build or adjust a plan toward a health goal, using your own insights as context. A training plan that knows your recovery markers. An eating plan that knows your cholesterol and blood sugar. A sleep plan that knows what your labs say about your iron. The plan is only as good as the goal you give it, so give it one goal at a time, and check the plan against what your doctor has already told you before you follow any of it.

Track progress over time

Ask Muse to compare your latest results with the round before, and tell you what improved, what got worse, and what stayed the same. This is the use that turns one off lab reading into a habit: every new round of results becomes a check in on the plan you are following. Did the number you were working on actually move? Did anything else move with it? A plan you never measure is a plan you never know is working, and a chat comparison after each round of labs is the lightest way to measure it. Keep the comparison short: three numbers that moved, and what they might mean, and then stop.

Prepare for your doctor, not instead of your doctor

The best use of this connector is the one that ends in a real appointment. Ask Muse to turn your results into the three questions you should ask your doctor, in order, with the numbers behind each one. You walk into the appointment with a short list instead of a vague worry, and the doctor gets a patient who knows their own numbers. Our complete connectors list shows the other health connectors Muse can reach, including Withings for device data and Apple Health for activity and health metrics on iPhone, and our Google Calendar connector guide covers booking the appointment itself once you know what to ask.

5. What it cannot do

The limits are honest and worth knowing before you lean on the connector.

• It cannot diagnose you. Muse is not a doctor. It can explain a result in plain words, but it cannot tell you what is wrong with you, and it must never be the thing that decides whether a symptom matters. Anything that worries you goes to a clinician, not to a chat.
• It cannot prescribe or change treatment. It cannot tell you to start, stop, or change a medicine, a supplement, or a therapy. That decision belongs to you and your doctor, with your full history in front of them. A plan Muse builds is a draft to discuss, never an instruction to follow.
• It cannot guarantee it read a result right. A summary can misread a number, a unit, or a trend. Check anything that touches your health against your Function summary before you act on it, and never quote a number to your doctor from a chat summary without checking the original first.
• It cannot see health data outside Function. Muse sees the Function account you connected, nothing more. Results from other labs, other apps, and other doctors are invisible to it, unless you connect those sources too. A partial picture can mislead, so tell Muse what it cannot see before you ask it to judge the whole picture.
• It cannot replace your clinician reviewed summaries. Function gives you clinician reviewed summaries for a reason: a human clinician has looked at your results. Muse reads those summaries, but it does not replace them. If Muse and your summary disagree, trust the summary and ask your doctor.

6. What to watch for

As of October 25, 2026, Function Health is a settled connector: live since September 8, still counted among the 54 connectors in the web app in the October 15 in app check, with no October update to how it works. The thing to watch is not the connector changing. It is how you use it. Health data is the most sensitive data you will ever hand an agent, and the habits around it matter more than the setup.

Never act on a health answer without checking it. If Muse says a number is fine, check the Function summary. If Muse says a number is worth asking about, check the Function summary, and then ask your doctor. The failure mode to avoid is treating a chat answer as a medical opinion. It is not one. It is a reading aid, and the person who decides what a result means is a clinician who knows your history.

Check your data controls. In Muse settings, open Data controls and look at the switch for helping improve the models. If you do not want your chats used for training, turn it off before you start asking Muse about your lab results. That switch controls whether Meta may use your chats to train future models. Think twice before you connect health data you would not want read aloud, and remember that you can disconnect Function Health at any time in Settings under Connectors. Disconnecting stops Muse from exchanging data with the connector, though information Muse previously used might still remain in your memories and conversation history.

Finally, keep the scope small on purpose. The health uses that work best through this connector are the ones you can describe in one sentence and check in two minutes: one number, one trend, one question for your doctor. When a health question needs a paragraph to explain, that question belongs in a doctor's office, not in a chat. Use Muse to prepare for the appointment, and let the appointment do the deciding.

Get started

Frequently asked questions

What does the Function Health connector do in Muse?

It brings your Function lab results and clinician reviewed health summaries into Muse. Muse uses them as context to help you understand your health, build plans around your goals, and track progress over time.

Can Muse diagnose me from my lab results?

No. Muse is not a doctor. It can explain a result in plain words and help you prepare questions for your doctor, but it cannot diagnose, prescribe, or tell you to start or stop any treatment.

Do I need a Function membership to use the connector?

Yes. The connector reads your Function account, so you need to be a Function member with lab results already in your account. A connector with no results behind it has nothing to show you.

Is my health data used to train Muse?

There is a Data controls switch in Muse settings for helping improve the models. If you do not want your chats used for training, turn it off before you start asking Muse about your lab results.

Can I use the Function Health connector outside the United States, Canada, and Japan?

Not yet. Muse itself has launched in the United States, Canada, and Japan only. Join the waitlist and we will send you a fresh code on launch day in your country.

What happens if I disconnect Function Health?

Muse stops exchanging data with the connector at once. Your lab results stay in your Function account. Information Muse previously used might still remain in your memories and conversation history, and you can ask Muse to forget those interactions.

Built with Muse is an independent community project. Not affiliated with or endorsed by Meta.