Platform Examples

Examples serve as both educational resources and practical starting points for building conversational AI applications. Each example demonstrates a specific use case, integration pattern, or application type, complete with pre-configured settings and best practices baked in.

Browsing Available Examples

The platform maintains a curated library of examples covering various integration types, industries, and use cases:

GET /api/v1/platform/example/list

http

Each example includes:

  • id: Unique slug identifier for the example
  • name: Descriptive title of the example
  • description: Overview of what the example demonstrates
  • type: The integration or application type (blueprint, project, widget, slack, discord, etc.)
  • tags: Searchable tags for discovering relevant examples

Example Types

Examples are categorized by their integration or deployment type:

  • blueprint: Complete agent configurations that can be deployed across channels
  • project: Full SDK code examples demonstrating integration patterns
  • widget: Web-embeddable chat widgets with specific styling and behavior
  • slack: Slack bot configurations and integration patterns
  • discord: Discord bot templates and community management examples
  • whatsapp: WhatsApp Business integration examples
  • messenger: Facebook Messenger bot configurations
  • telegram: Telegram bot templates
  • twilio: SMS and voice-based conversational examples
  • email: Email-based conversational automation
  • trigger: Event-driven automation patterns
{
  "id": "customer-support-assistant",
"name": "Customer Support Assistant",
"description": "AI assistant for handling common customer inquiries",
"type": "blueprint",
"tags": ["support", "customer-service", "helpdesk"],
"createdAt": 1700000000000,
  "updatedAt": 1700000000000
}

javascript

Using Examples as Templates

Examples are designed to be starting points that you can customize for your specific needs. The list endpoint provides lightweight metadata to help you browse and discover available examples. Common use cases include:

  • Browsing all available example templates
  • Discovering examples by type or category
  • Building example selection interfaces
  • Syncing example metadata to external systems

Each example represents tested, working configurations that demonstrate best practices for specific use cases, integration patterns, or industries.

Searching for Relevant Examples

When you have a specific use case or need in mind, semantic search helps you find the most relevant examples quickly. The search uses AI-powered similarity matching to understand the intent behind your query and return examples that best match your needs, even if they don't contain exact keyword matches.

To search for examples based on a natural language description:

POST /api/v1/platform/example/search
Content-Type: application/json

{
  "search": "chatbot for e-commerce customer support with order tracking"
}

http

The search parameter accepts natural language queries describing what you're trying to build or the problem you're trying to solve. The optional take parameter limits the number of results returned (1-100, default is 10).

The more specific your query, the more relevant the results will be. For example:

  • "sales assistant that qualifies leads and schedules meetings"
  • "discord bot for community moderation with automated warnings"
  • "multilingual customer service agent with FAQ integration"

Understanding Search Results

Search results are ordered by relevance, with the most semantically similar examples appearing first. The response structure matches the list endpoint, returning the same detailed information about each matching example. The search algorithm considers:

  • Example descriptions and use case documentation
  • Configured abilities and integrations
  • Keywords and categorical tags
  • Typical conversation patterns and behaviors

The take parameter controls how many results are returned (1-100, default 10). The similarity threshold is set to 0, meaning all examples are considered, but they're ranked by relevance score.

{
  "items": [
    {
      "id": "ecommerce-support-agent",
      "name": "E-commerce Support Agent",
      "description": "Handles order tracking and product questions",
      "type": "blueprint",
      "tags": ["ecommerce", "orders", "support"],
      "link": "https://chatbotkit.com/examples/ecommerce-support-agent",
      "createdAt": 1700000000000,
      "updatedAt": 1700000000000
    }
  ]
}

javascript

Best Practices for Searching

For optimal search results:

  • Be specific about your use case and requirements
  • Mention key features or integrations you need
  • Include industry or domain context if relevant
  • Use natural language rather than trying to match keywords
  • Adjust the take parameter based on how many alternatives you want to review

If your initial search doesn't return relevant results, try rephrasing your query with different terms or breaking down your requirements into simpler concepts. You can also browse the full example list to discover categories and use cases you might not have considered.

Cloning Examples to Your Account

Once you've found an example that fits your needs, you can clone it to your account as a starting point for customization. Cloning creates copies of all resources defined in the example, including bots, datasets, integrations, and any dependencies, properly configured and ready to use.

To clone an example:

POST /api/v1/platform/example/{exampleId}/clone
Content-Type: application/json

{}

http

Replace {exampleId} with the ID (slug) of the example you want to clone. For instance, to clone the customer support assistant example:

POST /api/v1/platform/example/customer-support-assistant/clone
Content-Type: application/json

{}

http

What Gets Cloned

The cloning process intelligently handles both simple and complex examples:

Blueprint Examples: Creates a complete agent configuration with all settings, personality traits, and model selections preserved.

Integration Examples: Sets up the necessary resources for the specific integration type (bot, widget configuration, channel settings) with all dependencies properly linked.

Complex Examples: For examples with multiple interconnected resources (datasets, skills, integrations), the system performs topological sorting to ensure dependencies are created in the correct order, then rewires all references to point to your newly created resources.

Clone Response Structure

The response indicates which resources were created during the cloning process:

{
  "resources": {
    "blueprint": [
      {
        "id": "clr9x8f3k000008l8e5j9h2m4",
        "name": "Customer Support Assistant",
        "description": "Cloned from example"
      }
    ],
    "dataset": [
      {
        "id": "clr9x8f3k000108l8g7h3k5n1",
        "name": "Product Knowledge Base"
      }
    ]
  }
}

javascript

The resources object contains arrays grouped by resource type. Each created resource includes its ID, name, and description, allowing you to immediately access and further customize the cloned resources.

Post-Clone Customization

After cloning, the newly created resources are completely independent from the original example. You can:

  • Modify configurations without affecting the template
  • Add or remove abilities and integrations
  • Customize conversation flows and responses
  • Adjust personality traits and behavior
  • Connect to your own data sources and APIs

Important: The cloning process may take several seconds for complex examples with many resources and dependencies. The operation has a 60-second timeout limit. If cloning fails, check that all dependencies and prerequisites are met (such as required secret types or subscription plan features).

Note: Some examples may require specific integrations or abilities that need additional configuration after cloning, such as API keys or OAuth authentication. Review the cloned resource configurations to complete any necessary setup steps.

Go to cbk.ai