Creating Your First Agent

A step-by-step guide to building your first AI agent

Let’s build your first AI agent in 5 minutes! Follow along step by step.

Diagram

Step-by-Step Guide

Install PraisonAI
Open your terminal and run:
bash
```
pip install praisonaiagents
```
Set Your API Key
Set your OpenAI API key:
bash
```
export OPENAI_API_KEY=your_key_here
```
Note
Get your API key from platform.openai.com

Create Your Agent File

Create a new file called my_agent.py:

python

from praisonaiagents import Agent

# Create your first agent
agent = Agent(
    name="ResearchAssistant",
    instructions="You are a helpful research assistant. Explain things simply and clearly."
)

# Ask it a question
agent.start("Explain how solar panels work in simple terms")

Run Your Agent
bash
```
python my_agent.py
```
You should see the agent’s response explaining solar panels!

How It Works

Diagram

Step	What Happens
Instructions	Tell the agent who it is and how to behave
Question	Your message or task
LLM	AI model processes everything
Response	Agent returns the answer

Try Different Agents

Coding Helper

python

from praisonaiagents import Agent

coder = Agent(
    name="CodeHelper",
    instructions="You are a Python expert. Explain code simply and provide working examples."
)

coder.start("How do I read a CSV file in Python?")

Creative Writer

python

from praisonaiagents import Agent

writer = Agent(
    name="StoryWriter",
    instructions="You write short, engaging stories with vivid descriptions."
)

writer.start("Write a short story about a robot learning to paint")

Math Tutor

python

from praisonaiagents import Agent

tutor = Agent(
    name="MathTutor",
    instructions="You explain math concepts step by step. Use simple language."
)

tutor.start("Explain how percentages work")

Complete Example

python

from praisonaiagents import Agent

# Create a research assistant
assistant = Agent(
    name="ResearchAssistant",
    instructions="""You are a helpful research assistant.

When answering:
- Be concise and clear
- Use bullet points when helpful
- Explain complex topics simply
- Admit when you don't know something"""
)

# Ask multiple questions
assistant.start("What are the benefits of exercise?")
assistant.start("How does the internet work?")
assistant.start("Explain climate change in simple terms")