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

# Mintlify

> Import documentation from Mintlify-powered sites into AGO

Import documentation from Mintlify-powered documentation sites into AGO. This connector uses web scraping to extract content, so no API key is required.

## Overview

| Feature            | Details                                 |
| ------------------ | --------------------------------------- |
| **Authentication** | None required                           |
| **Content Type**   | Developer documentation                 |
| **Method**         | Web scraping                            |
| **Hierarchy**      | From navigation config or URL structure |

## Prerequisites

* A publicly accessible Mintlify documentation site
* The base URL of the documentation (e.g., `docs.example.com`)

## Configuration

### Required Fields

| Field    | Description                        | Example            |
| -------- | ---------------------------------- | ------------------ |
| `domain` | Base URL of the Mintlify docs site | `docs.example.com` |

### Optional Fields

| Field              | Description                       | Example                       |
| ------------------ | --------------------------------- | ----------------------------- |
| `include_patterns` | URL patterns to include (fnmatch) | `["/api/*", "/guides/*"]`     |
| `exclude_patterns` | URL patterns to exclude (fnmatch) | `["/changelog/*", "/blog/*"]` |

## Setup Steps

1. Navigate to **Knowledge** → **Sources**
2. Click **Create Source**
3. Select **Mintlify** as the source type
4. Enter the documentation domain
5. (Optional) Configure include/exclude patterns
6. Click **Save**
7. Click **Sync** to start importing

## URL Pattern Filtering

Use fnmatch-style patterns to control what gets imported:

### Include Patterns

Only import pages matching these patterns:

```json theme={null}
{
  "include_patterns": [
    "/api/*",
    "/guides/*",
    "/tutorials/*"
  ]
}
```

### Exclude Patterns

Skip pages matching these patterns:

```json theme={null}
{
  "exclude_patterns": [
    "/changelog/*",
    "/blog/*",
    "/internal/*"
  ]
}
```

### Pattern Examples

| Pattern                   | Matches                                 |
| ------------------------- | --------------------------------------- |
| `/api/*`                  | `/api/authentication`, `/api/endpoints` |
| `/guides/getting-started` | Exact match only                        |
| `*/v2/*`                  | Any path containing `/v2/`              |

## How It Works

1. **Discovery**: Fetches the sitemap to find all pages
2. **Filtering**: Applies include/exclude patterns
3. **Hierarchy**: Uses the site's navigation config (docs.json or mint.json) to build the sidebar structure. If no config is available, infers hierarchy from URL paths
4. **Extraction**: Fetches markdown content from each page
5. **Storage**: Saves documents with metadata and links to the original pages

## Limitations

* Only works with publicly accessible sites
* May be affected by rate limiting
* Complex JavaScript-rendered content may not extract fully
* Respects robots.txt directives

## Troubleshooting

<AccordionGroup>
  <Accordion title="Pages Not Found">
    * Verify the domain is correct and accessible
    * Check if the site requires authentication
    * Ensure pages aren't blocked by robots.txt
  </Accordion>

  <Accordion title="Missing Content">
    * Some dynamic content may not be captured
    * Check if content is loaded via JavaScript
    * Review exclude patterns aren't too broad
  </Accordion>

  <Accordion title="Sync Takes Too Long">
    * Large sites may take time to crawl
    * Consider using include patterns to limit scope
    * Check for redirect loops
  </Accordion>
</AccordionGroup>

## Related Documentation

* [Knowledge Connectors](./knowledge-connectors) - All available connectors
* [Knowledge Source Management](../features/knowledge-source-management) - Source configuration