match-payments

Match invoices against bank transactions and report payment status by confidence level.

Updated Mar 19, 2026

Search Tags:#bookkeeping #finance #confidence scoring #dry-run #invoice matching #bank transactions #payment reconciliation

Ready to run

npx skills add https://github.com/karimStekelenburg/k3m-marketplace --skill match-payments-karimstekelenburg

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: match-payments
Source: https://github.com/karimStekelenburg/k3m-marketplace/tree/main/plugins/personal-finance/skills/match-payments
Command: npx skills add https://github.com/karimStekelenburg/k3m-marketplace --skill match-payments-karimstekelenburg

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Reconciling invoices against bank transactions manually is slow and error-prone, especially when tracking which invoices have been paid and which still need follow-up. ## Core Features & Use Cases - Automated Payment Matching: Runs a matching engine that pairs invoices with bank transactions using IBAN, amount, date, and vendor name signals. - Confidence-Based Review: Classifies matches as confirmed, high, probable, review, or unmatched, and requires explicit human confirmation for ambiguous review-level matches. - Dry-Run Mode: Preview what would be matched without writing anything to the database. - Use Case: A freelancer with dozens of outstanding invoices runs the matcher monthly to see which clients have paid, confirm borderline matches, and identify unpaid invoices eligible for QR payment code generation. ## Quick Start Ask the assistant to match my invoices against bank transactions using the finance data directory, optionally with a dry run first.

Frequently Asked Questions about match-payments

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I match invoices against bank transactions automatically?▼

Run the matching engine with your finance data directory, and it pairs invoices with transactions using IBAN, exact or near-exact amounts, dates, and fuzzy vendor names. Results are grouped by confidence level from confirmed to unmatched.

How do I preview payment matches without changing my database?▼

Pass the --dry-run flag to see what would be matched without writing anything to the database. The summary table still shows all confidence levels so you can verify results before committing.

What happens with uncertain invoice-payment matches?▼

Matches at review confidence level require explicit human confirmation. For each one, the invoice details and candidate transaction are shown, and you answer yes, no, or skip before anything is written.

What matching criteria determine payment confidence levels?▼

Confirmed requires IBAN plus exact amount, high allows amount within €0.05, probable uses amount, date, and fuzzy vendor name, and review relies on amount plus description keywords. Anything below these thresholds stays unmatched.

What can I do with invoices that remain unmatched?▼

Unmatched invoices are listed with vendor, amount, due date, and IBAN availability. Those with a known IBAN are candidates for QR payment code generation using the generate-qr skill.