Invoice Sorting and Archival Assistant

Problem addressed

Reimbursement workflows often break down less from missing invoices than from scattered files, inconsistent names, and manual amount or category checks. Invoice Sorting and Archival Assistant targets local directories: it turns subfolders containing PDF invoices into an archival-ready state, reducing manual reading, copying, tabulation, and packaging.

How it works

The skill follows a fixed pipeline:

• Directory scan: on first run it asks for a root directory and only processes subfolders under that root, reducing out-of-scope access.
• Valid filtering: it processes folders that contain at least one PDF invoice and skips folders without PDFs.
• Content extraction: it reads invoice numbers, total amounts including tax, and expense keywords from PDFs for renaming and classification.
• Sheet generation: it creates 发票统计表.xlsx in each valid subfolder, with serial number, invoice number, expense name, total amount, and a summary row.
• File archiving: it renames PDFs to {invoice_number}.pdf, then packages them as ZIP files based on single-folder or multi-folder strategies.

Classification uses keyword priority rules for categories such as transport, meals, lodging, office, and printing; unmatched invoices fall into Other and are worth a quick review before export.

Limits and notes

It is best suited to local folders already grouped by trip, vendor, or expense batch, rather than automatically pulling files from email, cloud drives, or chat. PDF parsing depends on readable text; scanned, image-only, or oddly formatted PDFs may fail extraction. Verify amounts and invoice numbers, especially for railway tickets where the e-ticket number differs from the official invoice number.