CLI Documentation

Installation

npm install -g @chatbotkit/cli
# or
yarn global add @chatbotkit/cli

shell

Environment Setup

Before using the CLI, set up your environment variables:

export CHATBOTKIT_SECRET="your-api-secret"
# Optional: For partner API access
export RUNAS_USERID="optional-user-id"

shell

Command Reference

Basic Commands

# Get help
chatbotkit --help

# Check version
chatbotkit --version

shell

Bot Management

# Create a new bot
chatbotkit create bot --name "CustomerSupport" --model "gpt-4"

# List all bots
chatbotkit list bots

# Get bot details
chatbotkit get bot <bot-id>

# Update bot
chatbotkit update bot <bot-id> --name "NewName"

# Delete bot
chatbotkit delete bot <bot-id>

shell

Dataset Management

# Create a dataset
chatbotkit create dataset --name "Training Data" --visibility private

# List datasets
chatbotkit list datasets

# Upload data to dataset
chatbotkit dataset upload <dataset-id> --file data.json

# Delete dataset
chatbotkit delete dataset <dataset-id>

shell

Integration Commands

# Set up website integration
chatbotkit create integration sitemap \\
  --url "<https://example.com/sitemap.xml>" \\
  --name "Website Crawler"

# Create widget integration
chatbotkit create integration widget \\
  --name "Chat Widget" \\
  --theme light \\
  --position right

# List integrations
chatbotkit list integrations

shell

Interactive Input

# The CLI will prompt for confirmation
chatbotkit delete bot <bot-id> --confirm

# Prompt for input if not provided
chatbotkit create bot

shell

Configuration Files

You can use YAML files for complex configurations:

# config.yaml
name: MyBot
type: bot
config:
  model: gpt-4
  temperature: 0.7
  backstory: "A helpful assistant"

yaml

# Apply configuration
chatbotkit apply -f config.yaml

shell

Common Usage Examples

Setting up a Complete Bot

# 1. Create a new bot
chatbotkit create bot \\
  --name "Support Assistant" \\
  --model "gpt-4" \\
  --backstory "Customer support specialist"

# 2. Create and upload training data
chatbotkit create dataset --name "Support FAQ"
chatbotkit dataset upload <dataset-id> --file faq.json

# 3. Link dataset to bot
chatbotkit bot link <bot-id> --dataset <dataset-id>

shell

Managing Website Integration

# Start website crawling
chatbotkit integration sitemap create \\
  --url "<https://example.com/sitemap.xml>" \\
  --name "Website Data" \\
  --crawl-depth 2

# Check crawl status
chatbotkit integration sitemap status <integration-id>

shell

Deployment Management

# Deploy a configuration
chatbotkit deploy -f solution.yaml

# Check deployment status
chatbotkit deploy status <deployment-id>

# Rollback deployment
chatbotkit deploy rollback <deployment-id>

shell

Error Handling

The CLI provides clear error messages:

# Example of handling missing credentials
$ chatbotkit list bots
Error: CHATBOTKIT_SECRET environment variable not set
Please set your API secret using:
export CHATBOTKIT_SECRET="your-api-secret"

# Example of invalid resource ID
$ chatbotkit get bot invalid-id
Error: Bot not found: invalid-id

shell

Best Practices

1. Always use -confirm for destructive operations
2. Use configuration files for complex setups
3. Check operation status after async operations
4. Use environment variables for credentials
5. Regularly backup configurations

Would you like me to expand on any specific commands or add more usage examples?