How to use Claude to plan meals and shopping lists
The author used Claude Code to maintain recipes, retrieve USDA nutrition information, suggest weekly combinations, and produce categorized shopping lists and daily nutrient reports.
Prompt to try
Editorial adaptation, not the source author’s exact prompt. Replace bracketed inputs and review tool permissions first.
Organize [recipe photos, PDFs and text] into [meal directory]. Apply my [portion and ingredient preferences], retrieve available nutrition data through [USDA API access], and propose [weekly meals]. Generate a categorized shopping list and daily nutrient summary, flagging missing data for my review.
Reach people exploring AI agents
See placements →What happened
The author reports generating weekly shopping lists and detailed nutrient reports, while using time previously spent on planning and lists for further analysis.
Reported by richardblythman on Hacker NewsHow the agent did it
- Added recipes from cookbook photos, PDFs and websites to a directory.
- Modified recipe quantities, ingredients and preparation details.
- Retrieved and stored recipe nutrition information through the USDA API.
Where a person stepped in
The author explicitly stayed in the loop, guided analysis, adjusted recipes and manually filled their online shopping cart.
A person guided the agent closely as it worked.
The author described using Claude Code to build a meal library and plan weekly food shopping. They added recipes to a directory from cookbook photos, PDFs, or websites. They also modified recipes to fit their needs, such as doubling quantities for lunch leftovers, replacing difficult-to-find ingredients, or adding preparation details. The agent retrieved nutritional information through the USDA API and stored it with the recipes.
For weekly planning, the author could choose meals from the library or let the agent suggest combinations. Claude then generated a shopping list organized into categories such as dairy and vegetables, along with detailed reports of daily protein and micronutrients. The author still manually filled an online shopping cart using that list.
The author also used meal-plan analysis while considering a supplement purchase and reported receiving suggested ingredient changes. Separately, they tracked digestive symptoms and eczema and described an initial hypothesis from two weeks of records, rather than a confirmed cause. They emphasized remaining involved and said the agent was not good at data analysis without them. Who it's for: people who want an editable recipe collection, weekly shopping lists, and nutrition summaries they can review themselves.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: Hacker News · richardblythman
Frequently asked questions
How were recipes added and adapted?
The author added cookbook photos or copied recipes from PDFs and websites into a directory, then adjusted quantities, ingredients, and cooking details as needed.
Did Claude fill the online shopping cart?
No. Claude generated a categorized shopping list, but the author said they still manually filled the online shopping cart.
How involved was the author in analysis?
The author said they remained very much in the loop and that the agent was not good at data analysis without their involvement.
More details
Remaining steps
- Selected weekly meals or requested suggested combinations.
- Generated a categorized shopping list and daily nutrition reports.
- Manually filled the online shopping cart using the generated list.
About this account
The agent retrieved nutrition information, suggested meal combinations and generated lists and reports with ongoing user involvement.
Platform: Hacker News. Attribution: https://news.ycombinator.com/item?id=46499028. Prompt: Suggested (written by Agents In Use, not the author's). Sourced by the Agents In Use pipeline.
Credit: richardblythman on Hacker News
Source published 5 Jan 2026
Source checked 14 Oct 2026
Tools and topics
Claude Code · USDA API · Recipe directory
Meal planning · Shopping lists · Nutrition reports
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.