An intelligent agent built with LangChain that can answer questions by searching the web using SerpAPI and processing results with OpenAI's GPT model.
- Python 3.9+
- OpenAI API key
- SerpAPI key
- Clone the repository and create a virtual environment:
python -m venv .venv
source .venv/bin/activate # On Windows use: .venv\Scripts\activate- Install dependencies:
pip install -r requirements.txtSet up your API keys using one of these methods:
For Unix/Linux/macOS:
export OPENAI_API_KEY='your-openai-api-key'
export SERPAPI_API_KEY='your-serpapi-api-key'For Windows (PowerShell):
$env:OPENAI_API_KEY='your-openai-api-key'
$env:SERPAPI_API_KEY='your-serpapi-api-key'Create a .env file in the project root:
OPENAI_API_KEY=your-openai-api-key
SERPAPI_API_KEY=your-serpapi-api-key
Run the agent:
python web_agent.pyThe agent will process two example questions:
- Current weather in London
- Information about the French president
Example output:
Question: What is the current weather in London?
==================================================
Thought: I need to search for current weather information in London
Action: web_search
Action Input: current weather London
Observation: [Search results...]
Final Answer: [Weather details...]
Question: Who is the current president of France? What is their height in centimeters?
==================================================
...
- 🔍 Web search capabilities using SerpAPI
- 🤖 OpenAI GPT-3.5 Turbo integration
- 📝 Detailed thought process logging
- ⚡ Efficient error handling
- 🛠️ Custom tool implementation
- 🔒 Secure API key management
- Ensure both API keys are properly set
- Check internet connectivity for web searches
- Verify Python version compatibility
- Make sure all dependencies are correctly installed
Key packages:
- langchain and related packages
- openai
- python-dotenv
- google-search-results (SerpAPI)
For full list, see requirements.txt