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

# Scenarios

> Build interactive conversation flows with the visual scenario editor — guide users through forms, choices, and multi-step processes.

Scenarios let you create structured conversation flows instead of open-ended chat. Use them to collect information, guide users through processes, or handle specific topics with a defined sequence of steps.

## How scenarios work

A scenario is triggered when the AI detects a matching condition in a user's message. Once triggered, the conversation follows a defined flow of nodes until it reaches an end node. The user can exit a scenario at any time using exit keywords.

## Creating a scenario

<Steps>
  <Step title="Open the scenario editor">
    Go to your agent's **Scenarios** tab and click **Create Scenario**. Give it a name and description.
  </Step>

  <Step title="Configure the trigger">
    Every scenario starts with a **Start node**. Choose how it gets triggered:

    | Trigger type | How it works                                                                                                                                                |
    | ------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------- |
    | **Keyword**  | Triggered when the user's message contains one of the specified keywords. Add one or more keywords.                                                         |
    | **Intent**   | Triggered when the AI detects the described intent in the user's message. Write a natural language description (e.g., "user wants to book an appointment"). |
    | **Always**   | Triggered for every conversation. Use sparingly — typically for onboarding flows.                                                                           |
  </Step>

  <Step title="Add nodes">
    Drag nodes onto the canvas and connect them to build your flow. See the node types below.
  </Step>

  <Step title="Connect nodes">
    Draw edges between nodes to define the flow. For question nodes with choices, you can set conditions on edges to create branching paths.
  </Step>

  <Step title="Save and enable">
    Click **Save** to store the scenario. Toggle **Enabled** to activate it for your agent.
  </Step>
</Steps>

## Node types

### Message node

Sends a text message to the user. You can optionally include a button.

* **Message text** — the message content
* **Show button** (optional) — display a clickable button with custom text and URL

### Question node

Asks the user a question and stores their answer in a variable.

* **Question text** — what to ask
* **Variable name** — where to store the answer (used in later nodes)
* **Input type**:
  * **Text** — free-text input with optional validation (email, phone, or number)
  * **Choice** — multiple-choice options (requires at least 2 choices)
* **Validation** — for text input: none, email, phone, or number
* **Required** — whether the user must answer
* **Confirmation** — optionally ask the user to confirm their answer

### Form node

Displays a multi-field form for collecting structured data.

* **Pre-form message** — text shown before the form appears
* **Form title** — heading of the form
* **Form fields** — add input fields (text, email, phone, textarea) and file uploads
* **Submit button text** — label for the submit button

### Delay node

Pauses the conversation for a specified duration.

* **Duration** — wait time in milliseconds
* **Show typing indicator** — display a typing animation during the wait

### End node

Concludes the scenario with an optional action.

* **End message** — final message to the user
* **Action on end**:
  * **None** — just end the scenario
  * **Save submission** — save all collected data
  * **Trigger webhook** — send collected data to an external URL
  * **Show button** — display a button with a link

## Branching

Connect a question node with choices to different paths using **edge conditions**:

| Condition    | Description                                            |
| ------------ | ------------------------------------------------------ |
| **Default**  | Fallback path if no other condition matches            |
| **Equals**   | Matches when the answer exactly equals a value         |
| **Contains** | Matches when the answer contains a value               |
| **Choice**   | Matches when the user selects a specific choice option |

This lets you create different flows based on user responses — for example, routing a support request differently based on the category selected.

## Scenario settings

* **Exit keywords** — words that let the user leave the scenario (e.g., "cancel", "exit", "stop")
* **Exit message** — message shown when the user exits
* **Save on exit** — whether to save partial data when the user exits
* **Priority** — determines which scenario triggers first when multiple scenarios match (higher number = higher priority)
* **Enabled** — toggle to activate or deactivate the scenario

## Viewing submissions

Go to your scenario's **Submissions** tab to see all data collected through the scenario. Each submission includes the form data, timestamps, and the conversation it came from.

## Validation rules

The scenario editor validates your flow before saving:

* Must have exactly **1 Start node**
* Must have at least **1 End node**
* Start node must have an outgoing connection
* Keyword triggers must have at least 1 keyword
* Intent triggers must have a description
* Choice questions must have at least 2 options
* Form nodes must have at least 1 field

<Note>
  Each scenario trigger costs **1 credit** when activated. The scenario itself doesn't cost credits beyond the initial trigger and any chat messages within it.
</Note>