Invoice Organizer

Problem

Reimbursement folders often mix invoices, airline reimbursement receipts, ride-hailing trip records, and settlement statements across PDF, OFD, and ZIP files. Manually checking invoice numbers, dates, and amounts is error-prone, especially when duplicate invoices or non-invoice documents must be separated before submission.

How it works

The skill scans a target directory, identifies document types, and extracts key fields such as invoice number, issue date, buyer, seller, item, amount, tax rate, tax amount, and tax-inclusive total. Airline receipts can include additional fields such as flight number and origin/destination. Ride-hailing trip records are explicitly marked as non-invoices. Duplicate invoices are detected by invoice number and highlighted in the generated workbook.

For monthly organization, it generates an Excel workbook with detailed records, duplicate checks, and rollups by month, quarter, year, and expense category. It can also archive files by major category, subcategory, and company entity. Ambiguous invoices are saved to _pending_invoices.json, then confirmed in two steps for major and subcategory.

Caveats

Trip records are not invoices and should be reported separately. Tax-inclusive totals should come from source documents, not recalculation. Scanned PDFs and legacy stamped invoices may need manual review. --month targets a month folder and should not filter files inside that folder by issue date.