wehub-resource-sync
90 lines
3.1 KiB
Python
90 lines
3.1 KiB
Python
"""
|
|
Lesson 1: Basic Agent Creation
|
|
|
|
This script demonstrates the fundamentals of creating a simple AI agent using the AWS Strands SDK.
|
|
|
|
We will build a weather assistant that can:
|
|
1. Understand a natural language query about weather.
|
|
2. Use the `http_request` tool to fetch data from the National Weather Service API.
|
|
3. Synthesize the data into a human-readable response.
|
|
"""
|
|
|
|
import os
|
|
from dotenv import load_dotenv
|
|
from strands import Agent
|
|
from strands.models.litellm import LiteLLMModel
|
|
from strands_tools import http_request
|
|
|
|
# Load environment variables from a .env file
|
|
load_dotenv()
|
|
|
|
# Define a detailed system prompt to guide the agent's behavior
|
|
WEATHER_SYSTEM_PROMPT = """You are a friendly and helpful weather assistant with HTTP capabilities.
|
|
|
|
Your primary function is to provide accurate weather forecasts for locations in the United States by using the National Weather Service API.
|
|
|
|
Follow these steps to fulfill a user's request:
|
|
1. First, if you don't have grid coordinates, use the points API endpoint to get them.
|
|
- For latitude and longitude: https://api.weather.gov/points/{latitude},{longitude}
|
|
- For a US zipcode: https://api.weather.gov/points/{zipcode}
|
|
2. The points API will return a `forecast` URL. Use this URL to make a second HTTP request to get the actual weather forecast.
|
|
3. Process the forecast data and present it to the user in a clear, easy-to-understand format.
|
|
|
|
When displaying your response:
|
|
- Highlight key information like temperature, precipitation, and any weather alerts.
|
|
- Explain technical terms in simple language.
|
|
- If you encounter an error, apologize and explain that you couldn't retrieve the weather information.
|
|
"""
|
|
|
|
|
|
def create_weather_agent() -> Agent:
|
|
"""
|
|
Creates and configures a weather-focused agent.
|
|
|
|
Returns:
|
|
An Agent instance configured with a model, system prompt, and tools.
|
|
"""
|
|
# Configure the language model (LLM) that will power the agent.
|
|
# We use LiteLLMModel to connect to a provider like Nebius, OpenAI, or Anthropic.
|
|
model = LiteLLMModel(
|
|
client_args={
|
|
"api_key": os.getenv("NEBIUS_API_KEY"),
|
|
},
|
|
model_id="nebius/deepseek-ai/DeepSeek-V3-0324",
|
|
params={
|
|
"max_tokens": 1500,
|
|
"temperature": 0.7,
|
|
},
|
|
)
|
|
|
|
# Create the agent instance.
|
|
# The agent is the core component that orchestrates the LLM, tools, and system prompt.
|
|
weather_agent = Agent(
|
|
system_prompt=WEATHER_SYSTEM_PROMPT,
|
|
tools=[http_request], # Grant the agent the ability to make HTTP requests.
|
|
model=model,
|
|
)
|
|
return weather_agent
|
|
|
|
|
|
def main():
|
|
"""
|
|
Main function to run the weather agent.
|
|
"""
|
|
# Create the weather agent
|
|
weather_agent = create_weather_agent()
|
|
|
|
# Define a user query
|
|
user_query = "Compare the temperature in New York, NY and Chicago, IL this weekend."
|
|
|
|
# Invoke the agent with the query and get the response
|
|
print(f"User Query: {user_query}\n")
|
|
response = weather_agent(user_query)
|
|
|
|
# Print the agent's final response
|
|
print("Weather Agent Response:")
|
|
print(response)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main() |