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

# LLMs Context

> Use llms.txt and llms-full.txt to give AI assistants complete context about the AgentRef documentation. Feed the entire docs into Claude, ChatGPT, Cursor, or any LLM-powered tool.

AgentRef publishes machine-readable documentation endpoints that follow the [llms.txt standard](https://llmstxt.org/). These endpoints let you feed the full AgentRef documentation into any AI assistant as context -- enabling more accurate answers, code generation, and integration help.

## Endpoints

| URL                                          | Description                                            | Best For                                   |
| -------------------------------------------- | ------------------------------------------------------ | ------------------------------------------ |
| `https://www.agentref.co/docs/llms.txt`      | Concise overview of all documentation pages with links | Quick orientation, smaller context windows |
| `https://www.agentref.co/docs/llms-full.txt` | Complete documentation content in plain Markdown       | Full context, comprehensive answers        |

## What is llms.txt?

The `llms.txt` standard provides a way for websites to publish their content in a format optimized for large language models. Instead of making an LLM crawl HTML pages, you give it a single plain-text file containing everything it needs.

* **`llms.txt`** -- A structured index of all documentation pages with titles, descriptions, and URLs. Think of it as a table of contents for AI.
* **`llms-full.txt`** -- The complete documentation rendered as plain Markdown in a single file. Every page, every code example, every table -- all in one response.

## Using with AI Assistants

### Claude

Paste the URL directly in a conversation:

```
Read https://www.agentref.co/docs/llms-full.txt and then help me set up
an affiliate program with 20% recurring commission.
```

Or use Claude Projects to add it as a knowledge source:

1. Create a new Claude Project
2. Add `https://www.agentref.co/docs/llms-full.txt` as a knowledge file
3. All conversations in the project will have full AgentRef context

### ChatGPT

Use the URL in a prompt or with Browse mode:

```
Fetch https://www.agentref.co/docs/llms-full.txt and use it as context.
I need to integrate AgentRef webhooks into my Express.js app.
```

### Cursor

Add the docs to your project context in `.cursorrules` or paste the URL:

```
@https://www.agentref.co/docs/llms-full.txt

Help me create a conversion tracking integration using the AgentRef Node SDK.
```

### Other LLM Tools

Any tool that accepts URL context or plain text can use these endpoints. The content is plain Markdown with no special formatting requirements.

## Tips for Optimal Context Usage

### Choose the right endpoint

* Use **`llms.txt`** when you only need the AI to know what documentation exists and where to find it. This is lighter on tokens and works well for navigation questions.
* Use **`llms-full.txt`** when you need the AI to have deep knowledge of AgentRef's APIs, SDKs, and features. This gives the best answers but uses more context window.

### Scope your questions

Even with full documentation context, you'll get better results by being specific:

```
# Good
"Using the AgentRef Node SDK, show me how to list pending payouts
and create a payout for each affiliate that meets the threshold."

# Less good
"Help me with AgentRef."
```

### Combine with code context

For integration tasks, give the AI both the AgentRef docs and your own code:

```
I'm building a Next.js app. Here's my current webhook handler:
[your code]

Read https://www.agentref.co/docs/llms-full.txt and help me verify
the webhook signature correctly.
```

### Use llms.txt for discovery

When you are not sure which part of the docs you need, start with `llms.txt`:

```
Read https://www.agentref.co/docs/llms.txt and tell me which pages
are relevant for setting up fraud detection.
```

Then follow up with the specific pages or switch to `llms-full.txt` for the complete context.

## Programmatic Access

You can fetch these endpoints from your code for automated workflows:

<CodeGroup>
  ```typescript Node.js theme={null}
  // Fetch the full docs as context for an LLM call
  const response = await fetch('https://www.agentref.co/docs/llms-full.txt');
  const docsContext = await response.text();

  // Use as system prompt context
  const messages = [
    {
      role: 'system',
      content: `You are an AgentRef integration assistant. Here is the full documentation:\n\n${docsContext}`
    },
    {
      role: 'user',
      content: 'How do I set up webhook signature verification?'
    }
  ];
  ```

  ```python Python theme={null}
  import httpx

  # Fetch the full docs as context for an LLM call
  response = httpx.get("https://www.agentref.co/docs/llms-full.txt")
  docs_context = response.text

  # Use as system prompt context
  messages = [
      {
          "role": "system",
          "content": f"You are an AgentRef integration assistant. Here is the full documentation:\n\n{docs_context}",
      },
      {
          "role": "user",
          "content": "How do I set up webhook signature verification?",
      },
  ]
  ```

  ```bash cURL theme={null}
  # Download for offline use
  curl -o agentref-docs.md https://www.agentref.co/docs/llms-full.txt

  # Quick preview
  curl -s https://www.agentref.co/docs/llms.txt | head -50
  ```
</CodeGroup>

## Content Updates

The llms.txt endpoints reflect the latest published documentation. When docs are updated, the endpoints serve the new content immediately -- no caching delay.

<Tip>
  Bookmark `https://www.agentref.co/docs/llms-full.txt` in your AI tools. Whenever you're building an AgentRef integration, load it as context for the most accurate, up-to-date assistance.
</Tip>