Xiaoyi Health Body Temperature Monitoring and Recording Smart Assistant

The pandemic is overbut thermometers remain? This open-source skill helps you make the most of your data

A few days agoI rummaged through a drawer and found several thermometers stashed at home — infrared forehead thermometersmercury onesdigital onesthe full setTo be honestduring the COVID periodtaking temperature every day became muscle memoryAlthough we no longer have to report dailytemperature monitoring remains a necessity for families with elderly or childrenBut did you know? Simply jotting down numbers is a waste; the temperature curve hides many bodily signals.

I tried using my phone's memo app to record temperaturesbut over time I found three problems: firstI would forget to record; secondthere was no intuitive way to see trends; thirdno alert if a fever occurredLaterwhile browsing GitHubI discovered this skill called XiaoYi Health Body Temperature Monitorwhich solved all the pain points at once.

In simple termsthis skill adds a temperature recorder to the Claw smart assistantIt allows you to manually or automatically input body temperatureand then automatically generates a line chartmaking it clear at a glance which days the temperature was high or lowThe best part is its built-in alert function — when the temperature exceeds 37.3°C (you can adjust the threshold)a notification is pushedSet a warning line before sleepand you'll know immediately if your child has a feverno more getting up in the middle of the night to feel their forehead.

Installation Command is Simple

npx skills add <repository_url> --skill xiaoyi-health/body-temperature

Ohand before installationmake sure you have the Node.js andenvironments set upIf you find the command line cumbersomeyou can also download the ZIP package and deploy manually in three steps: unzipnpm installregisterThe tutorial is all on GitHub.

Not Fancybut Truly Helps You Relax

This skill features a "six-layer architecture," which sounds impressive but actually separates data storagechart renderingnotification pushand configuration managementmaking it easy to modify the codeAfter allit's open-source — if you want to add voice announcements or integrate with a Xiaomi bandyou can just modify it yourselfI tried the basic functionsand the recording interface looks like this:

• Quick Record: Input temperature valueselect body site (oralarmpitear temperatureetc.)choose time (now or custom)and save.
• Family Profile: Add multiple memberseach with independent datano interferenceSeparately view data for elderly and children at home — super practical.
• Trend Chart: Supports dailyweeklyand monthly temperature curveswith mouse hover showing specific valuesLook at the curve to know if it's stable or feverish.
• Alert Settings: Global threshold and individual member thresholdssupporting multiple notification methods (system notificationemailWebhook)I set an alert at 37.5°CLast time my child had a slight feverthe push notification alerted me before I even noticed.
• Data Export: Export as CSV or JSONconvenient for showing to doctorsLarge hospitals now accept electronic temperature records.

After using it for two weeksI discovered a hidden feature: combined with historical datait can tell you "Your recent average temperature is 0.2°C higher than last week" or "Your morning temperature is always low." These little tips are quite helpful for understanding your own physical condition.

Pitfall Record + Tips

To be honestduring the first installationI got stuck at the "npx skills add" stepgetting errors that the repository couldn't be foundLater I found out it was because I hadn't added an SSH key to my GitHub accountWellif you encounter network issuesI suggest using the ZIP manual installation:

1. Go to the repository's main page on GitHub and click "Code" -> "Download ZIP"
2. Unzip to yourskills folder
3. Run npx skills register/skills/xiaoyi-health/body-temperature
4. Restartdone

Additionallythere are a few points to note:

• Body temperature data is sensitive health information. If you deploy it on a public serverremember to add authenticationThis skill does not enable authentication by default; it's fine for local usebut for external accessit's recommended to add a password.
• Automatic measurement currently only supports manual input. For those who have a Bluetooth thermometeryou can write your own script to interfaceI noticed there is a /plugins directory in the repositorywhich should have reserved interfaces.
• For multi-user sharing, it's best to create different devices for each personor use 's multi-user moduleOtherwisethe data will be mixed together.

Finallylet me share my usage habit: measure once after waking up in the morningtwo hours after lunchand before going to bedand record into the skillAfter a weekI basically grasp my temperature patternWomen can also use it to help estimate ovulation (temperature slightly rises by 0.3-0.5°C)AnywayI think it's better than those paid apps — after allit's open-sourceand the data is all in your own handsso no worries about privacy leakage.

Alrightif you also have many thermometers gathering dust at homewhy not let this skill help them serve their purpose? Go ahead and deploy it! By the wayif you encounter errors or have new ideasyou can submit an Issue on GitHub — the author is quite active.