Integrate with EBANX using LLMs

Overview

Use large language models (LLMs) to streamline your EBANX integration workflow. Whether you're exploring our API capabilities or setting up your first payment flow, these AI-friendly resources will help you get started faster and more effectively.

AI-Friendly Resources

llms.txt — Documentation Index for AI Agents

We maintain an llms.txt file that helps LLMs and AI agents discover our documentation. It follows the llms.txt standard and includes:

• LLM Agent Instructions — key concepts, authentication, payment lifecycle, and critical rules embedded at the top of the file.
• Full page index — every documentation page linked with a description.
• Quick links — getting started guide, API reference, and error codes.

Markdown Page Versions

Every documentation page is also available as a clean .md file. Simply append .md to any page URL:

HTML page	Markdown version
docs.ebanx.com/docs/payments/get-started	docs.ebanx.com/docs/payments/get-started.md
docs.ebanx.com/docs/payments/guides/accept-payments/api/brazil/pix	docs.ebanx.com/docs/payments/guides/accept-payments/api/brazil/pix.md

These markdown files are generated at build time with rendered component content — no JSX tags, no JavaScript, just clean text that LLMs can parse accurately.

EBANX Integration Prompt

The EBANX Sandbox Integration Prompt is a ready-to-use system prompt for AI assistants. It helps AI tools generate sandbox-ready integration code, walk you through API usage, and provide step-by-step guidance for payments, refunds, payouts, and more.

Vibe Coding with EBANX

"Vibe coding" means describing what you want in natural language and letting an AI assistant write the code. Here's how to get the best results with EBANX APIs.

Quick Start Prompts

Copy and paste these into ChatGPT, Claude, Cursor, or your preferred AI assistant:

Create a Pix payment (Brazil):

Using the EBANX Direct API (sandbox), create a Pix payment for R$50.00 in Brazil.
Use endpoint POST https://sandbox.ebanx.com/ws/direct with integration_key.
The customer's CPF is 853.513.468-93. Return the pix.qr_code_value from the response.
Refer to: https://docs.ebanx.com/docs/payments/guides/accept-payments/api/brazil/pix.md

Accept a credit card payment:

Using the EBANX Direct API (sandbox), tokenize a credit card with EBANX.js
and create a payment for $25.00 USD in Mexico.
Use https://sandbox.ebanx.com/ws/direct.
Important: never send raw card numbers to the API — always tokenize first.
Refer to: https://docs.ebanx.com/docs/payments/guides/accept-payments/api/credit-cards.md

Process a refund:

Using the EBANX API (sandbox), refund a payment with hash "abc123".
Use POST https://sandbox.ebanx.com/ws/refund with the integration_key and payment hash.
Refer to: https://docs.ebanx.com/docs/payments/guides/after-payments/refund-payment.md

Check payment status:

Using the EBANX API (sandbox), query the status of payment hash "abc123".
Use GET https://sandbox.ebanx.com/ws/query with integration_key and hash parameters.
Payment statuses: OP (Open) → PE (Pending) → CO (Confirmed) or CA (Cancelled).
Refer to: https://docs.ebanx.com/docs/payments/guides/after-payments/retrieve-payment.md

Create a payout (send money):

Using the EBANX Payouts API (sandbox), send R$100.00 via Pix to a recipient in Brazil.
Use POST https://sandbox.ebanx.com/ws/payout/create.
Refer to: https://docs.ebanx.com/docs/payout/payouts-overview.md

System Prompt Template

Use this as a system prompt when starting a new conversation about EBANX integration:

You are an expert EBANX integration developer. Use the following resources:
- API docs: https://docs.ebanx.com/llms.txt
- Integration prompt: https://docs.ebanx.com/ebanx-integration-prompt.txt
Key rules:
1. Use sandbox (sandbox.ebanx.com) for all examples unless told otherwise.
2. Never send raw card numbers — always tokenize with EBANX.js first.
3. Include document (CPF) for Brazilian payments.
4. payment_type_code is country-specific: pix (Brazil), oxxo (Mexico), pse (Colombia).
5. Check payment status with /ws/query — statuses are OP → PE → CO or CA.

Tips for AI Coding Assistants

Cursor / Windsurf / VS Code with Copilot:

• Add https://docs.ebanx.com/llms.txt to your project's AI context or .cursorrules file.
• Point the assistant to specific .md page URLs when asking about a payment method.

ChatGPT / Claude:

• Start conversations with the system prompt template above.
• Paste the integration prompt file content for comprehensive sandbox guidance.

Devin / Automated Agents:

• Reference llms.txt as the entry point for documentation discovery.
• The markdown page versions (.md suffix) are optimized for LLM consumption.

Why Use These Resources?

• Better LLM Performance: Plain text formats and markdown structures are easier for LLMs to parse and understand than HTML pages with JavaScript components.
• Improved Accuracy: By accessing clean documentation, structured data, and AI-optimized prompts, you get better and more reliable outputs from language models.
• Developer Efficiency: Skip repetitive tasks and quickly scaffold your integration using AI-generated templates and walkthroughs.
• Always Current: All AI-friendly resources are generated at build time, so they stay in sync with the documentation automatically.