Extracts invoice data from images and PDFs into categorized Markdown expense reports.
npx skills add https://github.com/wlzh/skills --skill wlzh-invoice-scanner
Please help me install this Agent Skill. Skill: wlzh-invoice-scanner Source: https://github.com/wlzh/skills/tree/main/wlzh-invoice-scanner Command: npx skills add https://github.com/wlzh/skills --skill wlzh-invoice-scanner
Manually sorting travel, accommodation, and meal invoices for expense reimbursement is slow and error-prone. This Skill scans a folder of invoice images, PDFs, or ZIP archives, recognizes each invoice type, extracts key fields, deduplicates receipts, and produces a categorized Markdown report with verified totals.
Ask the assistant to run the invoice scanner on your receipts folder, for example: scan the folder ./trip-invoices and generate a categorized expense report.
High-intent search queries and answers about installing and using this skill.
Point the Skill at a directory containing JPG, PNG, or PDF invoices, or a ZIP archive. It extracts each invoice's number, date, and amount, classifies them into five categories, and writes a Markdown report named after the folder.
Supported inputs are JPG, JPEG, PNG, PDF, and ZIP archives. Recognized invoice types include taxi and transit receipts, flight and train tickets, hotel invoices, refund fees, and general expenses like meals and office supplies.
Deduplication runs during recognition using invoice numbers, so identical invoices count once. Supporting documents like trip itineraries and hotel folios are treated as evidence for the formal invoice and are not billed separately.
After extraction, the Skill runs an independent Python calculation summing all invoice amounts and comparing the result against the report total within a 0.02 tolerance. If validation fails, the report is recalculated until it passes.
Yes. Including the keywords 招待费, 招待, or 客户 in the input path or prompt switches to entertainment mode, which outputs one row per invoice with invoice number, amount, and merchant, skipping totals and category statistics.
ZIP archives are automatically extracted, their contents flattened into the report directory, and the originals deleted. XML and OFD metadata files are removed before scanning, and a final ZIP of the cleaned folder is created.
evolving-machines-labSorts uploaded files by category and year, organizing them into folders.
claude-office-skillsOrganize and categorize business expenses into structured reimbursement reports.
ricableExtract invoice data, rename files, and sort receipts into tax-ready folders.
ClawLink-HQManage Zoho Invoice API integration for automated invoicing and billing.
