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

# AI Integration

> How to instantly build Blink integrations using AI agents like Cursor, Windsurf, or Claude.

Blink provides native, first-class support for AI coding agents. Instead of manually reading through our API reference to write fetch requests or define schemas, you can point your AI directly to our **Master Agent File**.

## The Master Agent File (`llms-full.txt`)

Our documentation automatically generates an AI-optimized file at `https://docs.blink.store/llms-full.txt`. This single file contains:

1. **Blink Architectural Golden Rules:** Strict instructions on when to use Embedded checkouts, how to handle deliverables securely, and webhook best practices.
2. **Complete OpenAPI Schemas:** Every endpoint, payload, and response format in our API.

Because modern AI models (like Claude 3.5 Sonnet and GPT-4o) have massive context windows, they can instantly read this entire file and write flawless integration code on the first try.

## How to use it in Cursor or Windsurf

If you are building your integration inside an AI IDE like Cursor or Windsurf, simply `@`-mention the URL in your chat prompt.

**Example Prompt:**

> `@https://docs.blink.store/llms-full.txt` I am building a Next.js application. Please write a server action that creates a Blink checkout session for my product `prod_123`, and then write the client-side code to trigger the embedded overlay checkout when the user clicks "Buy Now".

## How to use it in ChatGPT or Claude web

If you are using ChatGPT or Claude on the web, you can either:

1. Paste the `https://docs.blink.store/llms-full.txt` URL directly into your prompt.
2. Download the `.txt` file and attach it to your "Claude Project" or "Custom GPT" as persistent knowledge.