AI Agent Use Cases
Views —

AI Agent Use Cases

Added today

44All 189

Prompt to Prepare a Shopping Cart from a Recipe Photo

The author reports that Grok Bot receives a recipe photo, organizes the ingredients and quantities, and creates a shopping list by excluding ingredients already available at home.

Grok Bot
Life & Hobbies

1–20 of 189

Work & Research

Extracting Ledger Items from Receipt Photos

They made a receipt processing app using Claude Code. It extracts the store name, date, and amount, and sends uncertain items to a review screen.

Claude Code

How it works

  1. They built an app that extracts necessary items from receipt photos received via website or messenger.
  2. The extraction results are saved along with the original scanned text for error comparison, and uncertain values can be corrected directly by a person on the screen.

Results

They released a beta app with a searchable ledger. Receipts were mostly processed one at a time.

Notes

The process includes human verification of incorrectly read values.

Source

Reddit Original post ↗

Example prompt · Claude Code

Create an app that extracts ledger items from receipt photos and allows people to review them.

Input: Sample receipt [file], items to record [column list], currency [currency], execution environment [environment]. First, complete the workflow from photo upload to ledger saving, then expand to Telegram/Discord input after confirming necessary connections.

Extract supplier, date, total amount, tax, and items, and save them together with the original image. Do not fill in unreadable values with guesses; mark them for review. Implement a ledger screen where users can edit extracted values and search.

If you add foreign currency conversion, save the exchange rate and source for the receipt date, and preserve the pre-conversion amount. Use only the tax calculation rules I provide, and distinguish between category suggestions and confirmed values. Organize the necessary settings for database, file storage, and model connections.

Deliverables: App source code, data structure, execution instructions, and sample results. Test with blurry photos, crumpled receipts, missing currency, duplicate uploads, and sum mismatches, and show items that require human verification.