Does it support VAT and discount extraction?

Absolutely. It identifies Danish 'moms', line-item discounts ('rabat'), member offers, and multi-buy promotions to ensure the final total is accurate.

How does it handle different currency and number formats?

It specifically handles Danish number formats (using commas as decimal separators) and converts them to standard decimal formats for data storage.

Which Danish retailers are supported by this skill?

It is optimized for major grocery chains like Netto, Rema 1000, and Foetex, as well as utility providers like TDC, Oersted, and HOFOR.

Can it handle blurry receipt images?

Yes, the skill includes a confidence scoring system (0.0 to 1.0) that flags low-quality or unreadable scans for manual user review.

What happens if a vendor is not recognized?

The skill falls back to general extraction rules and suggests a learning workflow to create a custom parser for that specific vendor in the future.

Document Parsing for SmartSpender

Name: Document Parsing for SmartSpender
Author: peerjakobsen

bypeerjakobsen

•

Productivity & Workflow

Extracts structured line-item data from Danish receipts and PDF invoices to automate expense tracking and financial categorization.

This skill provides Claude with specialized rules for converting unstructured financial documents into high-fidelity data. Optimized for the Danish market, it can identify specific vendors like Netto, TDC, and Oersted, normalize vendor IDs, and accurately parse complex line items from both thermal grocery receipts and digital PDF invoices. It includes a comprehensive taxonomy for categorizing goods—from dairy and meat to utility consumption—and features built-in validation to ensure extracted totals match the document sum, providing a robust foundation for automated bookkeeping and personal finance management.

Key Features

01Granular product categorization taxonomy for groceries, utilities, and subscriptions

02Support for vendor-specific parsing rules (TDC, HOFOR, Oersted) and general fallback logic

03Automated vendor detection using email domains, logos, and bank transaction signals

04Advanced confidence scoring and validation to ensure data integrity and flag blurry scans

052 GitHub stars

06Detailed extraction of Danish grocery receipt patterns including discounts, VAT (moms), and bottle deposits (pant)

Use Cases

01Automating the extraction of service line items from monthly telecom and utility PDF invoices

02Converting physical grocery receipts into structured digital records for budget analysis

03Normalizing disparate receipt formats into a standardized schema for financial applications

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add peerjakobsen/smartspender document-parsing

For use in Claude.ai and ChatGPT

Download Skill