Weather Agent AI App

A chain-of-thought AI agent powered by Google Gemini that can fetch weather information and execute system commands through natural language queries.

Features

• 🤖 Chain-of-thought reasoning with structured planning
• 🌤️ Real-time weather information using wttr.in API
• 💻 System command execution capability
• 🔄 Automatic tool selection and execution
• ⚡ Rate limiting to prevent API quota exhaustion
• 📝 JSON-structured responses using Pydantic models

Prerequisites

• Python 3.8+
• Google Gemini API key

Screenshot

• 

Installation

1. Clone the repository:

git clone https://github.com/jaya6400/weather_agent_ai_app.git
cd weather_agent_ai_app

2. Create a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install google-genai python-dotenv requests pydantic

4. Create a .env file in the project root:

GEMINI_API_KEY=your_gemini_api_key_here

Usage

Run the agent:

python file.py

Example Queries

👉🏻 what is the weather in ranchi
🧠 Step 1: User needs Ranchi weather. Will call get_weather tool.
🛠️  Step 2: Calling get_weather(ranchi)
   ↳ Result: The weather in ranchi is Clear 15°C
🤖 Step 3: The current weather in Ranchi is Clear with a temperature of 15°C.

👉🏻 what is the weather in delhi
👉🏻 solve 2 + 3 * 5

How It Works

The agent follows a 3-step chain-of-thought process:

1. PLAN: Analyzes the user query and determines the action needed
2. TOOL: Executes the appropriate tool (get_weather or run_command)
3. OUTPUT: Returns the final result to the user

Available Tools

• get_weather(city: str): Fetches current weather information for a city
• run_command(cmd: str): Executes system commands (use with caution)

Rate Limiting

The agent includes built-in rate limiting:

• 1-second delay between API calls
• Automatic retry with 10-second wait on 429 errors
• Maximum 5 steps per query to prevent infinite loops

Project Structure

weather_agent_ai_app/
├── agent.py           # Main agent script
├── .env              # Environment variables (API keys)
├── .gitignore        # Git ignore file
└── README.md         # This file

Configuration

You can modify these settings in agent.py:

• max_steps = 5: Maximum iterations per query
• time.sleep(1): Delay between API calls (in seconds)
• model="gemini-2.0-flash-exp": Gemini model to use

Safety Notes

⚠️ Warning: The run_command tool can execute system commands. Use with caution and avoid running untrusted commands.

Troubleshooting

Rate Limit Errors (429):

• The agent automatically handles these with retries
• If persistent, increase the delay in time.sleep()

API Key Issues:

• Ensure your .env file contains a valid GEMINI_API_KEY
• Get your API key from Google AI Studio

License

MIT License

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Acknowledgments

• Powered by Google Gemini
• Weather data from wttr.in