Skip to main content
This guide walks you through creating and configuring a new Duckie agent from scratch.

​
Before You Start

For the best experience, have these ready before creating your agent:
PreparationWhy
Knowledge sources connectedSo your agent has information to work with
Guidelines createdSo your agent knows how to communicate
Runbook or WorkflowSo your agent knows what to do
You can create an agent without these and add them later, but your agent won’t be very useful until configured.

​
Creating a New Agent

1

Navigate to Agents

Go to Build → Agents in your Duckie dashboard.
2

Click Create Agent

Click the Create Agent button to open the creation drawer.
3

Enter Basic Information

Fill in the basics:
  • Name — A descriptive name (e.g., “Support Agent”, “Billing Bot”, “Technical Help”)
  • Description — What this agent does (optional but helpful for your team)
4

Select Starting Logic

Choose how the agent should operate:
  • Runbook — Select a runbook for flexible, AI-driven execution
  • Workflow — Select a workflow for deterministic, step-by-step execution
If you haven’t created either yet, you can select a default or skip and add later.
5

Save

Click Create to save your new agent. The agent detail view will open.

​
Configuring Your Agent

After creating, configure these sections in the agent detail view:

​
Knowledge

Select which knowledge the agent can access:
  1. Go to the Knowledge tab
  2. Choose specific sources to include, OR
  3. Select knowledge tags to filter available content
Start broad, then narrow. It’s often easier to give access to everything first, then restrict based on testing results.

​
Guidelines

Assign communication guidelines:
  1. Go to the Guidelines tab
  2. Select which guideline sections apply to this agent
  3. Guidelines are additive — select multiple sections to combine them

​
Guardrails

Set up safety constraints:
  1. Go to the Guardrails tab
  2. Select Escalation Rules that should trigger handoff to humans
  3. Select Restrictions that define hard limits on agent behavior

​
Classification

Configure how conversations are tagged:
  1. Go to the Classification tab
  2. Select which Categories the agent should assign
  3. Select which Attributes to extract

​
Testing Your Agent

Before deploying, always test in the playground:
1

Open Playground

Go to Test → Playground.
2

Select Your Agent

Choose your newly created agent from the dropdown.
3

Send Test Messages

Try various scenarios:
  • Common questions your customers ask
  • Edge cases and unusual requests
  • Messages that should trigger escalation
4

Review and Iterate

Check responses and execution steps. Adjust configuration as needed.

​
Best Practices

​
Naming

  • Use clear, descriptive names: “Billing Support Agent” not “Agent 1”
  • Include the purpose or team in the name
  • Add descriptions explaining what the agent handles

​
Configuration

  • Start simple — Begin with basic configuration and add complexity as needed
  • Test thoroughly — Use the playground before every deployment
  • Iterate — Refine based on real results, not assumptions

​
Deployment

  • Shadow first — Always start new agents in shadow mode
  • Review responses — Check shadow drafts before going live
  • Monitor closely — Watch performance metrics after going live

​
Common Patterns

​
General Support Agent

  • Broad knowledge access (all sources)
  • Friendly, helpful guidelines
  • Escalation rules for angry customers and complex issues
  • Categories: Billing, Technical, Account, General

​
Specialized Agent

  • Narrow knowledge access (tagged sources only)
  • Domain-specific guidelines
  • Strict guardrails for the domain
  • Specific categories and attributes

​
Triage Agent

  • Minimal knowledge (just enough to categorize)
  • Short, clarifying responses
  • Quick escalation to specialized agents or humans
  • Focus on classification over resolution

​
Next Steps

Agent Configuration

Deep dive into all settings

Test in Playground

Validate before deploying

Create a Deployment

Connect to a customer channel