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

# Behaviours

In the Behaviours page, you can define how your agent behaves during conversations, including lead capture flows and fallback or escalation handling.

<AccordionGroup>
  <Accordion title="Lead Collection">
    <Frame>
      <img src="https://mintcdn.com/techopsfwdslashai/b8VYn-z985HSkXyP/images/Lead-Collection-1.png?fit=max&auto=format&n=b8VYn-z985HSkXyP&q=85&s=0466ddacabf162eea1e198b7a6804312" alt="Lead Collection 1" width="2880" height="1620" data-path="images/Lead-Collection-1.png" />
    </Frame>

    Use this behaviour to set up when and how lead collection is triggered in agent conversations.

    ​Define conditions such as **Lead Capture Style**, or when leads are collected, either at the beginning or after a specific number of messages - configured under **Lead Capture Timing.**

    Choose which fields to capture from:

    * Name
    * Phone Number
    * Email ID​

    Click **Apply** to set up the behaviour, or **Remove** to disable it.
  </Accordion>

  <Accordion title="Fallback Responses">
    <Frame>
      <img src="https://mintcdn.com/techopsfwdslashai/b8VYn-z985HSkXyP/images/Fallback-&-Human-Escalation.png?fit=max&auto=format&n=b8VYn-z985HSkXyP&q=85&s=d21ce1a9bbf9e04ad0b2af7c80f80d99" alt="Fallback & Human Escalation" width="2880" height="1620" data-path="images/Fallback-&-Human-Escalation.png" />
    </Frame>

    Use this behavior to configure fallback rules for the agent.

    You can add a contact email to be triggered in case of a fallback. Additionally, you can define conditions by specifying trigger keywords and providing a fallback message.

    Click **Apply** to set up the behaviour, or **Remove** to disable it.
  </Accordion>

  <Accordion title="Human Escalation">
    <Frame>
      <img src="https://mintcdn.com/techopsfwdslashai/b8VYn-z985HSkXyP/images/Human-Escalation-1.png?fit=max&auto=format&n=b8VYn-z985HSkXyP&q=85&s=e55a7947c2a7468a6d3f1ce8dbf7ef48" alt="Human Escalation 1" width="2880" height="1620" data-path="images/Human-Escalation-1.png" />
    </Frame>

    Use this behaviour to set up Human Escalations.

    You can connect your Slack workspace and configure channels for the responses to go through when a human escalation is triggered. A human agent can join in on the live chat and respond to messages via the Slack channel.

    Click on **manage integration to configure channels** and **disconnect Slack** to disable ti
  </Accordion>
</AccordionGroup>

Now that your agent’s behavior is defined, you can get it to perform more complex tasks in the **Actions** page.