🤖🧠 Simple Agentic AI 🕵⚙️

Simple Agentic AI is a lightweight, curious little brain that knows when to stop thinking and start doing.

Instead of just answering questions, this project demonstrates how an agentic AI can decide which tool to use, call it, and return real-world information—just like a smart assistant should. No magic, no overengineering—just clean, understandable agent behavior.

Powered by the NousResearch/Hermes-2-Pro-Llama-3-8B LLM, this project keeps agent behavior transparent, understandable, and easy to learn from.

🧠 What makes it “agentic”?

Agentic AI isn’t just about generating text. It’s about:

• Understanding a user’s intent
• Choosing the right external tool
• Calling that tool
• Interpreting and presenting the result

Simple Agentic AI does exactly that—step by step—so you can see how modern AI agents actually work under the hood.

🛠️ Tools it can use

This agent knows when to reach outside itself and fetch live data:

• 🌤 WeatherAPI — for real-time weather conditions
• 🕒 RapidAPI (Date & Time) — for accurate current date and time across regions
• 🕌 Prayer Times API (AlAdhan) — for precise daily prayer times based on location

Each tool is called intentionally, only when needed—because a smart agent doesn’t guess, it checks.

🎯 Why this project exists

Simple Agentic AI is built to be:

• Educational — perfect for learning how tool-calling agents work
• Approachable — no massive frameworks or black-box behavior
• Practical — real APIs, real data, real use cases

Whether you’re exploring agent design, learning tool orchestration, or just curious how AI can interact with the real world, this project keeps things simple—and fun.

❓ About the APIs

Please follow these links to get your own API keys:

• WeatherAPI
• RapidAPI

Don't forget to include the keys to your environment:

export WEATHERAPI_API_KEY="YOUR_KEY_HERE"
export RAPIDAPI_KEY="YOUR_KEY_HERE"

💻 Example Usage

Clone this repository and install some packages:

git clone https://github.com/reshalfahsi/simple-agentic-ai
cd simple-agentic-ai
pip3 install -r requirements.txt

Create your own file or edit simple_agentic_ai.py:

from agent import SimpleAgenticAI


messages = [
    "How is the weather right now in Sendai? What time is it there?",
    "When is maghrib time on 26 February 2026 in Sendai?",
]


def main():
    model = SimpleAgenticAI()

    for message in messages:
        print("\nUser:")
        print(message)
        output = model.chat(message)
        print("\nAssistant:")
        print(output)


if __name__ == "__main__":
    main()

Save and run your Python file! For example:

python3 simple_agentic_ai.py

Happy tool calling, agent!