Quick Start

Get your first AI agent running in under 5 minutes.

Prerequisites

An AiQarus account at app.aiqarus.com
A configured organization and project

Step 1: Access the Agent Builder

Log in to your AiQarus dashboard
Navigate to Agents in the sidebar
Click Create Agent

Step 2: Configure Your Agent

The Agent Builder wizard guides you through configuration:

Basic Information

Field	Description	Example
Name	Human-readable agent name	”Customer Support Triage”
Description	What the agent does	”Categorizes and routes support tickets”
Project	Which project this belongs to	”Customer Success”

System Prompt

The system prompt defines your agent’s behavior. Write clear instructions:


You are a customer support triage agent. Your job is to:
1. Read incoming support tickets
2. Categorize them (billing, technical, general)
3. Assign priority (low, medium, high, urgent)
4. Route to the appropriate team

Always explain your reasoning for categorization and priority.

Safety Constraints

Set limits to prevent runaway behavior:

Constraint	Description	Recommended
Max Steps	Maximum TDAO iterations	10-50
Max Duration	Timeout in seconds	300 (5 min)
Token Limit	Max LLM tokens per run	10,000-50,000

Step 3: Activate the Agent

Review your configuration
Click Activate
Your agent is now ready to run

Once activated, an agent definition becomes immutable. To make changes, create a new version.

Step 4: Run Your Agent

Click Run on your agent
Provide input (structured form or natural language)
Watch the execution in real-time

Example Input


{
  "ticket": {
    "subject": "Cannot access my account",
    "body": "I've been locked out for 2 days. Urgent - I have a deadline tomorrow!",
    "customer_tier": "enterprise"
  }
}

Example Output


{
  "category": "technical",
  "priority": "urgent",
  "routing": "account_recovery_team",
  "reasoning": "Customer is locked out (technical issue). Enterprise tier with stated deadline makes this urgent. Routing to account recovery specialists."
}

Step 5: View the Audit Trail

Every decision your agent makes is recorded:

Go to the Runs tab
Click on your run
Expand each step to see the TDAO loop in action
Click Verify Chain to validate audit integrity

Next Steps

Learn about the TDAO execution loop
Configure human-in-the-loop for high-stakes decisions
Add integrations to connect external systems
Build multi-agent workflows