> ## 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.

# Quickstart

> Get your first AGO agent answering questions in 5 minutes

Get your first AGO agent answering questions in 5 minutes. For a deeper walkthrough covering prompt engineering, tools, testing, and monitoring, see [Build Your First Agent](./guides/getting-started-tutorial).

## Prerequisites

* Access to an AGO instance (SaaS or self-hosted)
* Admin or staff permissions
* At least one knowledge source to connect (Notion, Confluence, etc.)

***

<Steps>
  <Step title="Connect a Knowledge Source">
    <Frame caption="Knowledge Sources">
      <img src="https://mintcdn.com/ago/nVwBOovwlRSHxKDS/features/images/knowledge-sources.webp?fit=max&auto=format&n=nVwBOovwlRSHxKDS&q=85&s=b135d9f9cf33cc733f4bfd2d2ae6bcbb" alt="Knowledge Sources page showing connected sources with their content types, sync status, and actions" width="1280" height="800" data-path="features/images/knowledge-sources.webp" />
    </Frame>

    1. In the admin sidebar, click **Knowledge**, then select **Sources**
    2. Click **Add Source** and select your platform (e.g., Notion, Confluence)
    3. Authenticate and select the content to sync
    4. Click **Save**

    Wait for the sync indicator to turn green.

    <Tip>Start with a small subset of documents to test quickly. You can add more content later.</Tip>
  </Step>

  <Step title="Create an Agent">
    <Frame caption="AI Studio - Agents List">
      <img src="https://mintcdn.com/ago/nVwBOovwlRSHxKDS/features/images/ai-studio-agents-list.webp?fit=max&auto=format&n=nVwBOovwlRSHxKDS&q=85&s=23bc3a8ed0d83ba9eced4b70abbdddf2" alt="AI Studio showing a list of configured agents with their models, tools, and status" width="1280" height="800" data-path="features/images/ai-studio-agents-list.webp" />
    </Frame>

    1. In the admin sidebar, click **AI Studio**, then select **Agents**
    2. Click **Create Agent**
    3. Give your agent a **Name** (e.g., `support-agent`) and a **Display Name** (e.g., `Support Assistant`)
    4. Under **Knowledge Sources**, select the source you just connected
    5. Click **Create Agent**
  </Step>

  <Step title="Set Agent Permissions">
    1. In the agent builder, go to the **Agent Visibility** section
    2. Under **Permissions**, select the groups that should access this agent (e.g., `Default`)
    3. Click **Save**
  </Step>

  <Step title="Test Your Agent">
    1. Open the chat interface
    2. Ask a question related to your knowledge source
    3. Verify the response includes accurate information with source citations

    <Tip>Use the **Search Lab** (in AI Studio) to debug document retrieval if answers seem off.</Tip>
  </Step>

  <Step title="Deploy">
    Choose how users will interact with the agent:

    <CardGroup cols={2}>
      <Card title="Widget" icon="window" href="/features/widget-configuration-admin">
        Embed the chat widget on your website
      </Card>

      <Card title="Slack" icon="hashtag" href="/data-connector/slack-integration">
        Deploy in a Slack channel
      </Card>

      <Card title="API" icon="code" href="/api/public-api-v1">
        Send messages programmatically
      </Card>

      <Card title="SDK" icon="brackets-curly" href="/features/sdk-integration">
        Build a custom UI with the JavaScript SDK
      </Card>
    </CardGroup>
  </Step>
</Steps>

***

## What's Next?

Continue with [Build Your First Agent](./guides/getting-started-tutorial) to learn about prompt engineering, tools, simulation testing, and monitoring.

> **⚠️ WARNING**: Before production deployment, configure proper [authentication](./security/security) and review [access control](./users/access-control-rules) settings.