Agents Building Agents
Describe what you need in plain English. Our AI builds a fully configured agent in seconds using a 3-prompt pipeline.
The 3-Prompt Pipeline
Each user request passes through three specialized AI prompts, each handling a different aspect of agent creation.
Intent Analyzer
Understands what you're trying to build
Parses your natural language description and extracts key requirements: industry, use case, required capabilities, tone, and target audience.
// Input
"I need a booking assistant for my hair salon that can schedule appointments and remember customer preferences"
// Output
{
"industry": "beauty/salon",
"use_case": "appointment_booking",
"capabilities": ["scheduling", "memory", "preferences"],
"tone": "friendly, professional"
}Personality Designer
Crafts the agent's character and voice
Creates a detailed system prompt that defines the agent's personality, communication style, knowledge boundaries, and behavioral guidelines.
// Generated System Prompt
"You are Bella, a friendly and professional booking assistant for [Salon Name].
Your personality:
- Warm and welcoming, like greeting a regular customer
- Efficient but never rushed
- Remember and reference past appointments
You can help with:
- Scheduling haircuts, coloring, styling appointments
- Recommending services based on preferences
- Sending appointment reminders
Always confirm: service type, stylist preference, date/time."Configuration Builder
Generates the complete agent config
Produces the final agent configuration including name, slug, greeting message, suggested prompts, database schema, and integration settings.
// Final Agent Configuration
{
"name": "Bella - Salon Assistant",
"slug": "bella-salon-xyz123",
"greeting": "Hi! I'm Bella, your salon assistant. Ready to book your next appointment?",
"suggested_prompts": [
"Book a haircut",
"What services do you offer?",
"Cancel my appointment"
],
"capabilities": {
"booking": true,
"memory": true,
"notifications": true
}
}See It In Action
Example Generation
Alex - Support Assistant
Hi! I'm Alex, here to help with any questions about your account. What can I assist you with today?
Generation Stats
Pipeline Architecture
┌─────────────────────────────────────────────────────────────────────────────────┐
│ AGENTS BUILDING AGENTS PIPELINE │
└─────────────────────────────────────────────────────────────────────────────────┘
User Input: "I need a restaurant booking bot with menu recommendations"
│
▼
┌─────────────────────────────────────────────────────────────────────────────┐
│ PROMPT 1: INTENT ANALYZER │
│ │
│ Extract: │
│ • Industry: restaurant/hospitality │
│ • Primary function: reservations │
│ • Secondary: menu recommendations │
│ • Implied: hours, location, dietary filters │
└─────────────────────────────────────────────────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────────────────────────────────────┐
│ PROMPT 2: PERSONALITY DESIGNER │
│ │
│ Generate: │
│ • Name: "Marco" (Italian restaurant vibe) │
│ • Tone: Warm, knowledgeable about food │
│ • System prompt with personality traits │
│ • Conversation boundaries │
└─────────────────────────────────────────────────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────────────────────────────────────┐
│ PROMPT 3: CONFIGURATION BUILDER │
│ │
│ Output: │
│ • Complete agent config JSON │
│ • Database schema for reservations │
│ • Greeting message │
│ • Suggested user prompts │
│ • Integration webhooks │
└─────────────────────────────────────────────────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────────────────────────────────────┐
│ LIVE AGENT CREATED │
│ │
│ Marco - Restaurant Host │
│ "Buongiorno! I'm Marco, ready to help you reserve a table or explore │
│ our menu. How can I assist you today?" │
│ │
│ API Endpoint: /chat/marco-restaurant-abc123/message │
└─────────────────────────────────────────────────────────────────────────────┘
What Gets Generated
The pipeline creates everything needed for a production-ready agent:
System Prompt
Detailed personality, knowledge, and behavioral guidelines
Greeting Message
Engaging first message that sets the tone
Suggested Prompts
Quick-start buttons for common user actions
Database Schema
Tables for bookings, preferences, user data
API Reference
Build Agent
POST /api/smart-builder/build
// Request
{
"prompt": "I need a customer support agent for my e-commerce store"
}
// Response
{
"success": true,
"agent": {
"id": 42,
"name": "Sophie - E-commerce Support",
"slug": "sophie-ecom-x7k2m",
"greeting": "Hi there! I'm Sophie, ready to help with orders, returns, or product questions.",
"system_prompt": "You are Sophie, a friendly e-commerce support specialist...",
"suggested_prompts": [
"Track my order",
"Start a return",
"Product recommendations"
]
},
"chat_url": "/chat/sophie-ecom-x7k2m"
}Magic: One API call, one sentence, one fully configured AI agent ready to chat with your users.