Jithox logo

Jithox

Verified compliance checks for AI agents over MCP

Overview

About Jithox

Jithox is a set of verified compliance and payment-safety tools that AI agents call over the Model Context Protocol (MCP). Instead of letting an agent guess at a fact that carries real financial or legal risk, Jithox gives it an authoritative read-only check: verify an IBAN, confirm a Peppol e-invoicing participant, test whether a supplier bank-change request is safe, run an EU import or energy-label preflight, or screen a counterparty against sanctions lists. The problem it addresses is that language models invent plausible answers, and in finance, invoicing, and trade compliance a confident wrong answer can move money to the wrong account or clear a shipment that should have been stopped. Jithox is designed so a truthful negative - no match, not registered, not reachable - is a real, charged answer, which removes any incentive to fabricate a match.

Jithox stays deliberately read-only today. Its live tools look things up in official sources and change nothing, and the vendor is explicit that higher action levels (compare, prepare, approve, execute, reconcile) are not claimed until each is implemented and proven. The service connects to any MCP-speaking agent (Claude, ChatGPT and Codex, Gemini, Cursor, Goose, and plain HTTP or CLI agents) by adding a single endpoint, and several core tools - IBAN verification, Peppol readiness and participant lookup, and payment-change preflight - are free and need no account or key. The commercial model is pure usage: you pay per accepted call from one prepaid EUR balance, with no subscription, no renewal, and no overage. A machine-readable pricing contract lets an agent read the cost of a call before making it.

Pricing

Priced in another currency

Trial only
  • Free tools (IBAN, Peppol, payment preflight)$0/mo
  • E-Invoice / Import / Energy-Label preflight (EUR 0.01 per call)Not listed
  • Sanctions preflight (EUR 0.09 per call)Not listed

From the vendor pricing page, 2026-10-06

Details

Data from: Website•Updated: Oct 6, 2026
agent-frameworktask-automationautonomous-agentmcp