Logging food with AI

Record food in your own collections, using your chosen properties, presets and AI instructions.

---

Food uses the same Collections, Properties and Entries as anything else you track. Choose the fields you want to record and tell your AI how to use them. Collection names and nutritional fields are yours to choose.

Set up your tracking

Create a collection with the properties you need, or copy a suitable setup from the Library. Use a preset for a repeated meal. If you track ingredients separately, a preset can include sub-entries.

In your skills, describe any rules you want the AI to follow: which collection to use, which amounts to record, and whether to estimate missing nutritional values. Starter skills can be edited, disabled or deleted just like your own.

Ask to record something

In any AI chat, make your intent explicit. For example:

Log my lunch in my meals collection using my usual lunch preset.

The AI uses your existing collection properties and presets. It does not require collections named Food or Dish. Any estimates should be identified as estimates, and you can edit recorded values afterward.

Include photos

Use the + button to upload images or take a photo on your phone. You can also paste an image into the web composer. Several photos can accompany one request.

When an entry or preset is successfully logged, the app can save the photos to that entry's image property. The target collection needs an image property. Attaching a photo to ask a question is not itself a request to log an entry.

Check or undo the result

The action below the answer shows what was saved. Open the entry to check its values or use the action's revert control. A multi-step request can produce several actions; review each saved result.

The same collection and skill setup applies on web, desktop and mobile.