> ## Documentation Index
> Fetch the complete documentation index at: https://docs.moduluslabs.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Build with AI

> Scaffold a working Modulus integration by pointing an AI agent at these docs

## Overview

These docs are built to be read by AI coding agents as well as people. Every page has a **Copy page** control and an **Open in Claude / ChatGPT** action, and the full documentation is exposed in machine-readable form at `/llms.txt` (index) and `/llms-full.txt` (complete text). You can point Claude Code, Cursor, or any AI agent at the docs and have it scaffold a working integration.

<Steps>
  <Step title="Give the agent the docs">
    Share the relevant pages, or the `/llms-full.txt` URL, with your agent. For an Ecom integration, point it at the [Ecom Quickstart](/docs/ecom/quickstart) and the [API Reference](/api-reference/ecom/create-payment-link). Use the Copy or Open in Claude control on any page to hand over its exact contents.
  </Step>

  <Step title="Prompt it to build">
    Be specific about the environment and the money format. For example:

    ```
    Using the Modulus Ecom docs, scaffold a Node.js server that creates a
    payment link (POST /payment-links with X-API-Key and Idempotency-Key) and a
    minimal checkout page. Amounts are integers in PHP centavos. Use the sandbox
    base URL https://api.sbx.moduluslabs.io/ecom/v1.
    ```

    For an embedded card form instead of a hosted link, point the agent at the [JavaScript SDK](/docs/ecom/jssdk/introduction) and ask it to mount the card fields and call `confirmPayment`.
  </Step>

  <Step title="Test with sandbox cards">
    Run the integration and pay with a sandbox [test card](/docs/testing): challenge card `4000 0000 0000 2503`, OTP `1234`. No real money moves.
  </Step>
</Steps>

<Note>
  Keep secrets out of the agent's context. Provide keys through environment variables at runtime, not in the prompt. Secret keys (`sk_`) belong on the server only.
</Note>