Connect Google Gemini to your health data and log it all by chat.

Wellness Project는 다음과 연동돼요 Google Gemini as a custom app in Gemini Spark, Google’s personal intelligence surface. Tell Gemini what you ate, how you trained, or how you slept and it lands structured in the right tracker, then Gemini reads your trends back when you ask. It’s the same OAuth-protected MCP server Claude, ChatGPT, Grok, and Le Chat use, and there is no API key or Client ID to paste.

다운로드iPhone다운로드Android대신 웹 브라우저에서 로그인

다운로드iPhone 다운로드Android

무료 · 현재 iPhone 제공 · Apple Health, Fitbit 및 Oura 동기화

독립 보안 검증 완료

전송 및 저장 시 암호화돼요.건강 데이터는 절대 판매하지 않아요.언제든 모든 데이터를 내보내거나 삭제할 수 있어요.데이터 보호 방법

안에 들어있는 것

Gemini as the front-end for logging and tracking your health

로그 meals from Gemini

"Log lunch: chicken burrito bowl, rice, and guac." Gemini hands it to Wellness Project, which estimates the calories and macros and files it in your nutrition log. No label-scanning, no forms.

운동과 러닝 기록

"Log my push day: bench 4x6 at 195, OHP 4x8 at 115." Gemini structures the sets and reps; Wellness Project saves the session, keeps your volume history, and detects PRs.

수면, 체중 등 기록

Most modules Wellness Project tracks are callable from Gemini: sleep, body weight, supplements, recovery, wellbeing, runs, lab results. It becomes a general health log, not only a training tracker.

Read your data in Gemini

"how did training look this week?" or "is my HRV up?" and Gemini queries Wellness Project for the real figures from your own logs and synced wearables, instead of answering in generalities.에게 물어보세요

No API key, no Client ID

Our registration endpoint is open (Dynamic Client Registration, RFC 7591), so Gemini registers itself on first connect. You paste one URL and sign in; the Client ID and Client Secret fields, if your client shows them, stay blank.

Google is named, and you can cut it off

The authorization screen tells you the data will be handled by Google under your agreement with Google, and every connection shows up under Settings, Devices & Integrations, Connected AI assistants with a Disconnect button that revokes read and write access on the next request.

차근차근

Connect Gemini Spark in five steps

1
Gemini Spark 열기
Go to gemini.google.com/spark and sign in with the Google account you use for Gemini.
2
Connected apps 열기
In Spark, open Settings → Personal intelligence → Connected apps. This is where Spark lists the apps it is allowed to reach.
3
커스텀 앱 추가
Choose Add a custom app. Spark asks for the address of the app you want to connect.
4
MCP 서버 URL 붙여넣기
Enter wellnessproject.ai/api/mcp and follow the prompts. There is no Client ID or Client Secret to enter, because the connector registers itself.
5
로그인해서 승인하기
A Wellness Project sign-in and approval screen opens. Sign in with the same account you use in the app, approve, and Gemini loads the tool list. Then just say things like "log my workout" or "what did I eat yesterday?"

앱에서

Gemini can already talk about fitness. This is what makes it remember.

Gemini for a meal plan or a training block and you get one, but nothing persists: the next conversation starts from zero, and none of it becomes a trend. Wellness Project is the missing half. Connect it once as a Spark custom app and every meal, set, and night of sleep you mention is에게 물어보세요 기록되고 유지됨 in your own account, so it compounds into history you can ask about later.

The connection is a two-way MCP link. Gemini writes to your tracker when you log and reads from it when you ask, and what it reads includes everything already syncing in: Fitbit, Oura, Withings, Apple Health, and Google Health Connect, the Android hub Pixel and Samsung watches write to. It’s the same /api/mcp endpoint and the same tool set every other assistant reaches, so your data layer is identical no matter which one you prefer.

One thing to expect during setup: the authorization screen says Wellness Project doesn’t recognize the app and names the host that will receive your data. That notice is informational and expected here, because Google doesn’t publish a fixed callback host for its connectors. The generic protocol details are at Settings → Connect any MCP client, and the same server also works with Claude, ChatGPT, Grok, 그리고 Mistral Le Chat.

자주 묻는 질문

How do I connect Google Gemini to a fitness tracker?

In Gemini Spark (gemini.google.com/spark), open Settings → Personal intelligence → Connected apps → Add a custom app, enter wellnessproject.ai/api/mcp, and follow the prompts. A Wellness Project sign-in screen opens; approve it with the same account you use in the app and Gemini loads the tools.

Do I need an API key or a Client ID for Gemini?

No. Our registration endpoint is open, so a compliant client registers itself on first connect and there is no key to generate. If a field asks for a Client ID or Client Secret, leave it blank. Your account is identified when you sign in, over OAuth with PKCE.

Why does the authorization screen say Wellness Project doesn't recognize this app?

Because Google does not publish a fixed OAuth callback host for its connectors, so we cannot positively attribute the request to a specific product. The notice is informational, not a warning that something is wrong: it names the exact host that will receive your data so you can check it looks like Google before approving.

Can Gemini log my meals and workouts, or only read them?

Both. Gemini writes to your account when you log a meal, a workout, sleep, weight, or a supplement, and reads back the same data when you ask about a trend. Correcting and deleting entries works from the conversation too.

Does this work with my Android health data?

Yes. Anything already syncing into Wellness Project is readable from Gemini, including Google Health Connect, the Android hub that Pixel, Samsung Health, and most Android trackers write to, plus direct connections like Fitbit, Oura, and Withings.

Is my data shared with Google?

Only the slice Gemini pulls during your active conversation. We do not batch-export your data; the model reads the specific rows it asks for, when it asks, over a connection you can revoke at any time from Settings, Devices & Integrations.