Meal-prep recipe that hits your nutrition goals, plus shopping list

A 5-lunch chicken, rice and broccoli prep hit stand-in goals by USDA data (622 kcal, 63.5 g protein, 49.2 g carbs) with a complete shopping list; the Instacart order was not tested.

Score

10.0/10

claims that held up · score and verdict differ

Verdict

Partly reproduced

21

Access needed

L0 Public web only

Run 21 Oct 2026

Score and verdict differ. Every claim we tested held up; the Instacart order was not tested, so the verdict stays Partly.

2

claims scored

1 not tested

2 of 2

held up in full

1

sites and APIs touched

about 30 seconds

run time

Claim by claim

The recipe matches the nutrition goalscoreHeld up

USDA recompute: 622 kcal, 63.5 g protein, 49.2 g carbs per lunch, all within the goals.

Writes a shopping list with all the ingredientscoreHeld up

All 4 main ingredients listed in enough quantity; chicken margin only ~10 g.

Orders on Instacart and gets it deliveredcoreNot tested

Not tested (spends money).

Open a claim to see what we checked.

Access it needs
L0 Public web only
The original post's setup: L0 recipe + L4 (orders and pays on Instacart)
Run on
21 Oct 2026, 09:14 to 09:15 (UTC+8)
Run by
Grok Bot
Check again by
20 Nov 2026

What was claimed

'My Grok Bot "Chef" just created a meal prep recipe matching my nutrition goals, wrote a shopping list, created an order on instacart with all the ingredients, and had it delivered to my house.'

@JordanKeto on X. Original post

What we found

Lemon-garlic chicken, brown rice and broccoli bowls: 180 g roast chicken breast, 150 g cooked brown rice, 150 g broccoli, 10 g olive oil per lunch. USDA numbers: 622 kcal, 63.5 g protein, 49.2 g carbs per lunch, so all three goals are met (the agent's own estimate, ~620 kcal / ~60 g / ~50 g, was close). The shopping list covers all 5 lunches, but the chicken is tight: 1.25 kg raw bought vs about 1.24 kg needed. The Instacart order and delivery were not tested.

How we checked it

Programmatic: USDA FoodData Central (SR Legacy) values for each ingredient (fdcId 171477, 169704, 169967, 171413; raw 171077, 169703 for cooking yield).

What the agent did

  1. Wrote recipe + estimate + shopping list (saved before the check)
  2. USDA FDC API: 4 cooked foods + 2 raw foods for cooking yield, all 200
  3. Recomputed per-lunch kcal, protein, carbs
  4. Checked list quantities against 5 lunches (raw-to-cooked yield from USDA)

Sites and APIs touched: api.nal.usda.gov.

Where it went wrong

  • Not blind: we set the goals and wrote the recipe, so this is a best case.
  • Raw-to-cooked weight is the easy place to go wrong: the list only works because it converts cooked weights back to raw (chicken loses about a quarter).
  • Oil lost in the pan and real cooking yields vary; treat the macros as about +/-10%.

Try it yourself

Exact prompt we used
Create one meal-prep recipe for 5 lunches. Each lunch: 550-650 kcal, at least 40 g protein, at most 60 g carbs. Give gram weights per ingredient, per-lunch macros, and a shopping list for all 5.

Lowest-access way to try it

L0 for the recipe and list. Give exact goals and ask for gram weights and the data source for each number. Let it fill a cart if you like, but approve the order and total yourself (L4).

Run details

Status
Partly reproduced (V1)
Access level used
L0: No account access
Original post's setup
L0 recipe + L4 (orders and pays on Instacart)
Method
The post does not give the goals, so we set stand-in goals: 5 prepped lunches, each 550-650 kcal, at least 40 g protein, at most 60 g carbs. The agent wrote the recipe, per-lunch weights, its own macro estimate and a shopping list first; then we recomputed every number from USDA data.
Run by
Grok Bot (this assistant) on its own Linux computer with web search/fetch/shell/browser tools; this run: wrote the recipe, then checked it with Python against the USDA FoodData Central API (no browser, no order)
Agent in the original post
Grok Bot (per post)
Run date
2026-10-21 (UTC+8)
Time
About 30 seconds (09:14:52-09:15:23 UTC+8)
Cost
Not metered (subscription agent); no paid APIs, no purchases
Re-verify by
2026-11-20

How to cite

TriedAgents, "Meal-prep recipe that hits your nutrition goals, plus shopping list", partly reproduced, checked 2026-10-21, https://triedagents.com/cases/meal-prep-to-nutrition-goals/

Get one re-run case every weekday

One agent use case we actually re-ran, every weekday. Free.

We only store your email address. Every issue has a one-click unsubscribe link. Privacy