Log a meal from a description

Sends a freeform meal description to Claude for per-item calorie and protein estimates, respecting stated portions and excluding anything the user says they skipped, then stores it as a meal.

POST/api/nutrition/describe

Authorization

__Secure-better-auth.session_tokenAPI key · cookierequired

Better Auth session. Signing in (Google or email magic link) sets an HttpOnly session cookie that scopes every request to that account. The web app and CLI use this; agents use the OAuth 2.1 flow of the MCP server.

AuthorizationBearer token · headerrequired

Personal API key sent as Authorization: Bearer skcal_…. Create and scope keys in the app under your profile → API keys. Keys default to full access; a scoped key is limited to its granted resource:action permissions and cannot manage other keys.

Query parameters

datestring<date>

Day to log the described meal against (YYYY-MM-DD). Defaults to the server's current UTC date.

Request body

requiredapplication/json

The freeform meal description to analyze.

textstringrequired

What you ate, in your own words. State anything you skipped so it is excluded.

max length 2000

Responses

200

The meal description was analyzed and logged.

okbooleanrequired

Always true once analysis succeeds.

mealIdstring<uuid>required

Identifier of the created meal.

itemsAnalyzedItem[]required

Estimated food items.

Show properties

Array of AnalyzedItem

namestringrequired

Estimated food item name.

kcalintegerrequired

Estimated calories for the portion.

proteinGnumberrequired

Estimated protein in grams.

totalKcalintegerrequired

Summed calories across items.

totalProteinGintegerrequired

Summed protein in grams across items.

notestringrequired

The model's one-line assumptions note.

photoKeysstring[]required

R2 object keys of stored photos (empty for text descriptions).

400

The request body or parameters failed validation.

errorstringrequired

Human-readable failure reason.

detailstring

Optional extra context, present on upstream failures.

502

The upstream model call failed.

errorstringrequired

Human-readable failure reason.

detailstring

Optional extra context, present on upstream failures.

503

A required secret (model key or ingest token) is not configured.

errorstringrequired

Human-readable failure reason.

detailstring

Optional extra context, present on upstream failures.

Try it

Server

Base URLCustom base URLOverrides the selected server when set.

Authorization

API keyRemember on this deviceInclude my values in samples

Parameters

datequeryDay to log the described meal against (YYYY-MM-DD). Defaults to the server's current UTC date.

Bodyapplication/json

textstring*What you ate, in your own words. State anything you skipped so it is excluded.

Request

curl -X POST 'https://app.skcal.fit/api/nutrition/describe' \
  -H 'Cookie: __Secure-better-auth.session_token=YOUR_API_KEY' \
  -H 'Accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "text": "two scrambled eggs, a slice of sourdough, and black coffee"
}'

const response = await fetch("https://app.skcal.fit/api/nutrition/describe", {
  method: "POST",
  headers: {
    "Cookie": "__Secure-better-auth.session_token=YOUR_API_KEY",
    "Accept": "application/json",
    "Content-Type": "application/json"
  },
  body: "{\n  \"text\": \"two scrambled eggs, a slice of sourdough, and black coffee\"\n}"
});

import requests

response = requests.post(
    "https://app.skcal.fit/api/nutrition/describe",
    headers={
        "Cookie": "__Secure-better-auth.session_token=YOUR_API_KEY",
        "Accept": "application/json",
        "Content-Type": "application/json"
    },
    data="{\n  \"text\": \"two scrambled eggs, a slice of sourdough, and black coffee\"\n}",
)

Response

{
  "ok": true,
  "mealId": "3f8b1c2a-5d6e-4f70-8a1b-2c3d4e5f6071",
  "items": [
    {
      "name": "grilled chicken breast",
      "kcal": 284,
      "proteinG": 53.4
    },
    {
      "name": "toum",
      "kcal": 90,
      "proteinG": 0.4
    }
  ],
  "totalKcal": 374,
  "totalProteinG": 54,
  "note": "assumed a single chicken breast and one tablespoon of toum",
  "photoKeys": [
    "nick@mintlify.com/2026-06-29/3f8b1c2a-5d6e-4f70-8a1b-2c3d4e5f6071"
  ]
}

{
  "error": "weightKg must be 9-320 kg",
  "detail": "upstream model returned 500"
}

{
  "error": "weightKg must be 9-320 kg",
  "detail": "upstream model returned 500"
}

{
  "error": "weightKg must be 9-320 kg",
  "detail": "upstream model returned 500"
}