Use Claude to turn order history into dinner reminders
Lauren Morris struggled with dinner decisions after a day of meetings. Rather than building another recipe-planning app, she used Claude Code to retrieve her last thirty Instacart and DoorDash orders through Chrome and save them as files.
laurenlol.com
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Read my recent grocery and takeout orders, helping me identify five dinners I realistically make. Save patterns and exclusions, then build a skill that asks about this week's needs and generates a shopping list.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- Lauren Morris
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- Five personalized dinner patterns, a shopping-list skill and Telegram messages timed for shopping and weekday dinner decisions; no automatic ordering reported.
- Human involvement
- Morris selected patterns, wrote exclusions and chose dinners; scheduled jobs only delivered file content.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Retrieve order history and separate relevant household choices.
- Write reusable dinner patterns and exclusions into files.
- Generate shopping needs and schedule messages containing selected options.
Lauren Morris struggled with dinner decisions after a day of meetings. Rather than building another recipe-planning app, she used Claude Code to retrieve her last thirty Instacart and DoorDash orders through Chrome and save them as files.
She separated her own choices from orders for other household members and identified five workable dinner patterns. They used foods she actually ate, shelf-stable proteins and prepared ingredients that required little chopping. She wrote those patterns into one file and exclusions into another.
A Claude Code skill reads the files and asks two questions: how many dinners she wants to make and whether anything currently sounds unappealing. It produces a shopping list. The same setup can inspect a pantry screenshot to identify which dinner patterns are already supported and which items are missing.
Scheduled Python jobs on a Mac Mini send the shopping list through Telegram on Saturday morning, then dinner options Monday through Thursday at 3:30 p.m. Those jobs read Markdown and post messages; they are not fresh agent decisions or automatic grocery orders. Morris chooses what to eat and sometimes still orders takeout. The result is her own earlier decisions resurfacing when she has less energy to make them.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: laurenlol.com · Lauren Morris
Frequently asked questions
Why begin with order history?
Morris wanted patterns from food she actually ate, rather than another unfamiliar recipe library.
Were the reminders autonomous agent decisions?
No. Simple scheduled Python jobs read Markdown and posted the existing shopping list or dinner options to Telegram.
Who decided what to eat?
Morris made the planning choices and selected dinner from the delivered options, sometimes still choosing takeout.
More details
About this account
Claude extracted browser order history and supported a file-based dinner skill and reminder setup.
Firsthand source opened and read. Suggested prompt written by us from the reported workflow; no original task prompt is quoted.
Credit: Lauren Morris
Source published
Added to Agents In Use
Source checked
Tools and topics
Claude Code · Chrome · Telegram · Markdown · launchd
Dinner decisions · Order history · Telegram reminders
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.