Skip to content
PromptReady to runFinanceDocuments

Receipt Extractor

Written in-house · editorially graded

Every SkillsBank offering is written in-house — never scraped or crowd-sourced — and graded against a fixed editorial rubric in a separate grading step.

Dry-run test passed · Jun 15, 2026

Editorially graded · Jun 15, 2026

Task: Transform a batch of receipt photos into populated expense rows. The prompt must instruct an OCR model to read each image, extract the total amount and transaction date, infer merchant from the receipt header, and output a CSV with columns Date, Merchant, Amount, Category (auto‑filled as "Uncategorized"). Input: A list of image URLs or uploaded files. Output: CSV text ready to import into the daily expense tracker, plus a short note on any images that could not be parsed.

What you get

A ready-to-paste prompt you can drop into your AI.

Works in
Claude · ChatGPT · Gemini
Paste it in, fill the [brackets]
Any AI assistant
Works anywhere you can paste text

See it in action

A real input → output sample

Example Input
I've uploaded three receipt images — two good, one blurry. Please extract the expense rows.
Example Output
Receipt #,Date,Merchant,Amount,Currency,Category
1,2025-06-12,Le Bistro Paris,47.80,EUR,Meals & Entertainment
2,2025-06-13,Holiday Inn Express Lyon,135.00,EUR,Accommodation

Reader Note
1. Receipt 3 — unreadable image; row omitted.
2. Receipt 1 — Currency symbol was €; ISO code used EUR.
Complete

Goes well with

Related