GLP-1 medications work best when combined with a lifestyle program of diet and physical activity.
weight loss
GET /v1/public/health-fact?q=weight loss
Weight loss happens when the body uses more calories than it takes in.
General health information from federal public-domain sources — not medical advice, diagnosis, or treatment.
That was one real, public federal answer — and it's free. Receptivity scoring and the audience-safety verdict on every cue come with a free key.
The facts are public and free — that's the safety model. The free key unlocks the live cue loop — every cue scored for receptivity and checked safe for its audience. Clinical arc sequencing switches on with a paid stream.
Receptivity scoring
Every cue returns a receptivity score — when this patient is most likely to act.
Audience-safety verdict
Every message checked safe for who's receiving it.
The live cue loop
Create a patient stream, send a response token, get the next federal message — in your app.
One fact is the demo. The arc is the product.
The call above returns one fact. On a paid stream, Signal sequences the whole federal library into a per-episode arc — each message placed by day and phase, checked safe for the patient, and timed to when they are most likely to act. Here is the opening phase of a real clinical arc.
GLP-1 medications work alongside a lifestyle program of diet and physical activity.
The most common GLP-1 side effects are stomach-related: nausea, vomiting, diarrhea or constipation.
GLP-1 medications also slow the emptying of the stomach so food stays longer after meals.
Smaller meals work better than large meals when food stays in the stomach longer.
Side effects often improve as your body adjusts over the first few weeks.
This is a fixed illustrative slice. The live engine picks each next message from the full federal library — sequenced to the arc, timed to receptivity, and checked safe for the patient.
From one fact to the full engine
Ask
GET /v1/public/health-fact?q=…
One GET call, no key. Returns a real, federally-sourced fact, list, or tip — exactly what you are using right now.
Get a free key
Display modeA sandbox key in about five minutes — no PHI, no BAA. Render real federal messages in your own app.
Sequence
POST /v1/signal/streams/:id/cue
Send a patient's context. Get the next message — receptivity-scored, audience-safety-checked. Paid streams draw from the full federal library, sequenced to a clinical arc.
Five-minute integration
$ curl "https://api.uphealth.us/v1/public/health-fact?q=cholesterol" { "fact": { "kind": "fact", "body": "Having high blood cholesterol puts you at risk for heart disease, the leading cause of death in the U.S.", "agency": "NHLBI", "source_url": "https://www.nhlbi.nih.gov/health/blood-cholesterol", "uphealth_url": "https://uphealth.me/health/cholesterol-basics" }, "topic": "Cholesterol Basics", "disclaimer": "General health information from federal public-domain sources — not medical advice…", "attribution": "Source: NHLBI via Uphealth", "upgrade": { "signup_url": "https://uphealth.us/signup" } }
Call it from the OpenAI and Claude agent SDKs, the MCP server, or plain HTTP.
Or paste
https://mcp.uphealth.us/mcp
into Claude or ChatGPT as a custom connector —
MCP quickstart →.
Evaluating for your org? Your assistant can read the samples, check the sourcing, and write the memo — evaluate with your assistant →.
“We’ll just use Twilio.”
Twilio sends. Signal decides what to send, whether it is safe for this person, and what comes next.
“The content is free.”
It is — we source from the same CDC and NIH pages. You pay for the sequencing, the audience-safety verdict, and not maintaining freshness and linting yourself.
One key. The whole engine.
Building something for real? We pair with a few founding partners →