Stage an Instacart cart from repeat-buy patterns

An agent scans order history, predicts restocks, and loads the cart on a schedule while leaving checkout to the person.

Done with OpenClaw· Intermediate

What they did
They used OpenClaw to pull each day’s Instacart purchase records into a local SQLite file through a community CLI, keyed by Instacart item IDs. A SQL routine compares the spacing between repeat purchases and tags products as ready, likely soon, or not needed. Then, at 7:45am on Wednesday and Saturday, a scheduled job places the ready items in the cart by ID, and an iMessage follows at 8am.
How it went
Replacing browser-based clicking with direct item IDs raised accuracy from 11 of 14 products misplaced on a normal run to 100%. The process is deterministic and contains no LLM, and the user finishes the order manually because the CLI cannot submit it.
Tips
Direct IDs are the important part if you want the cart contents to match the history. This took roughly 20 minutes to set up, required an active Instacart session, and the timing was handled with macOS launchd.

Try it yourself with OpenClaw

Look through my [Instacart] order history and figure out which household items are probably ready to be bought again. At [day/time], use the exact item IDs to build a cart with those items, then send me a short text listing the items, quantities, and total. Stop before purchase and wait for my approval before any checkout or submission step.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on github.com ↗

1 more report of this

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.