🤖 AI Agent with Web Search & Weather Tools

An intelligent AI chatbot agent powered by LangChain that can search the web, provide weather information, and maintain conversation memory.

---

📋 Overview

This project provides a ReAct agent that can:

• Chat intelligently using Groq or OpenRouter LLMs
• Search the web using DuckDuckGo (no API key required)
• Get weather information from OpenWeatherMap
• Remember conversation context during the session
• Handle multiple tools dynamically

---

✨ Key Features

Feature	Details
LLM Support	Groq (free) or OpenRouter - choose your provider
Web Search	DuckDuckGo integration - no API key needed
Weather Tool	Real-time weather data from OpenWeatherMap
Conversation Memory	Remembers last 6 exchanges in session
ReAct Agent	Intelligent reasoning and action loop
Easy Config	All settings via .env file

---

🚀 Quick Start

1️⃣ Prerequisites

• Python 3.8 or higher
• An API key from Groq (free) or OpenRouter
• (Optional) OpenWeatherMap API key for weather tool

2️⃣ Installation

# Clone or download the project
cd basic-websearch-ai-agent

# Create virtual environment
python -m venv venv

# Activate virtual environment
# Windows
venv\Scripts\activate
# Mac/Linux
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

3️⃣ Configuration

Create a .env file in the project root:

# Choose ONE LLM provider:
# Option 1: Groq (Free)
GROQ_API_KEY=your-groq-api-key-here
GROQ_LLM_MODEL=llama3-8b-8192

# Option 2: OpenRouter
# OPENROUTER_API_KEY=your-openrouter-api-key-here
# OPENROUTER_LLM_MODEL=openai/gpt-3.5-turbo

# Optional: Weather tool
OPENWEATHER_API_KEY=your-openweather-api-key-here

# Model settings
TEMPERATURE=0.7

4️⃣ Run the Agent

python agent.py

---

💬 Usage Examples

🤖 Agent is ready! Type 'exit' to end the chat.

You: What's the current weather in New York?
Agent: The current weather in New York is partly cloudy. 
       The temperature is 18°C, but it feels like 15°C. 
       Humidity is at 65%, and the wind speed is 8 m/s.

You: Search for latest AI news
Agent: [Searches web and provides recent AI news articles]

You: Remember what I asked about weather?
Agent: [Uses memory to provide context-aware response]

You: exit
👋 Goodbye!

---

📁 Project Structure

basic-websearch-ai-agent/
├── agent.py              # Main agent execution
├── setting.py            # Pydantic configuration management
├── time_tool.py          # Get current time tool
├── requirements.txt      # Python dependencies
├── .env                  # Environment variables (create this)
├── .env.example          # Example environment file
├── .gitignore            # Git ignore rules
├── weather/
│   ├── logic.py          # Weather API integration
│   └── schema.py         # Input validation schema
└── README.md             # This file

---

🛠️ Tools Available

1. Web Search Tool

• Provider: DuckDuckGo
• No API key needed
• Returns recent web search results

2. Weather Tool

• Provider: OpenWeatherMap
• Requires: API key
• Returns: Temperature, humidity, wind speed, description

3. Time Tool

• Gets current date and time
• Useful for time-sensitive queries

---

🔧 Configuration Details

LLM Providers

Provider	Free Tier	Speed	Best For
Groq	✅ Yes	⚡ Very Fast	Best choice for beginners
OpenRouter	✅ Limited	🔄 Variable	Multiple model options

Environment Variables Explained

GROQ_API_KEY              # Your Groq API key
GROQ_LLM_MODEL            # Model name (e.g., llama3-8b-8192)
OPENROUTER_API_KEY        # Alternative to Groq
OPENROUTER_LLM_MODEL      # Model name for OpenRouter
OPENWEATHER_API_KEY       # Optional - for weather tool
TEMPERATURE               # 0.0 (deterministic) to 1.0 (creative)

---

🎯 How It Works

1. User Input → You type a question
2. Agent Reasoning → Agent decides which tool(s) to use
3. Tool Execution → Calls web search, weather, or time tool
4. Response Generation → LLM creates intelligent response
5. Memory Update → Stores conversation for context
6. Output → Agent responds to you

---

📝 Example Queries

• "What's the weather in London?"
• "Search for machine learning tutorials"
• "What time is it now?"
• "What are the latest technology trends?"
• "Tell me about climate change" (searches web + provides info)

---

⚠️ Troubleshooting

Issue	Solution
"No API key found"	Check .env file has GROQ_API_KEY or OPENROUTER_API_KEY
Weather tool fails	Ensure OPENWEATHER_API_KEY is set in .env
Slow responses	Check internet connection or LLM provider status
Memory errors	Restart the agent (memory resets between sessions)

---

🔐 Security Notes

• ✅ Never commit .env file to git (already in .gitignore)
• ✅ Keep API keys private
• ✅ Don't share .env file
• ✅ Use read-only API keys if provider allows

---

📚 Dependencies

Key libraries used:

• LangChain - Agent framework
• Groq/OpenAI - LLM providers
• DuckDuckGo - Web search
• Pydantic - Configuration & validation
• Requests - HTTP library

---

🤝 Contributing

Feel free to:

• Add new tools (calculator, code executor, etc.)
• Switch LLM providers
• Improve prompts
• Add persistent memory/database

---

📄 License

This project is open source. Use freely!

---

💡 Tips

1. Free Usage: Use Groq API (completely free with good limits)
2. Customize: Modify setting.py to add new environment variables
3. Extend: Add more tools in agent.py tools list
4. Debug: Set verbose=True in AgentExecutor to see agent reasoning

---

Last Updated: October 2025 | Status: Active Development