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

# Build a support agent with Clerk Chat

> Train, test, publish, and operate an AI support agent from one dashboard.

Clerk Chat helps you build a customer support agent from your existing content. Add knowledge, test replies, publish the agent, and review customer conversations from one dashboard.

<CardGroup cols={2}>
  <Card title="Create your first agent" icon="sparkles" href="/docs/quickstart">
    Go from a new workspace to a tested support agent.
  </Card>

  <Card title="Add knowledge" icon="database" href="/docs/guides/add-knowledge">
    Train your agent with websites, files, text, and question-and-answer pairs.
  </Card>

  <Card title="Publish your agent" icon="rocket" href="/docs/guides/deploy-agent">
    Add a chat widget to your site or publish a hosted help page.
  </Card>

  <Card title="Review conversations" icon="messages-square" href="/docs/guides/manage-conversations">
    Take over chats, reply to customers, and return control to the agent.
  </Card>
</CardGroup>

## How Clerk Chat works

Your agent uses three parts of your setup to answer customers:

* **Instructions**: define how the agent responds and which support policies it follows
* **Knowledge sources**: provide the product information used in answers
* **Deployment channel**: makes the agent available through a widget or hosted help page

Use the [quickstart](/docs/quickstart) to configure all three parts.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.