We'll build a weather assistant that can understand a question, fetch live data from an external API, and provide a helpful answer. This will introduce you to the core concepts of the Strands SDK:

Agent Creation: How to instantiate an agent.
Model Configuration: How to connect your agent to a large language model (LLM).
Tool Usage: How to give your agent abilities, like accessing the internet.

Key Concepts Explained

System Prompt: The WEATHER_SYSTEM_PROMPT is the agent's constitution. It's a detailed set of instructions that defines the agent's personality, its capabilities, and the exact steps it should follow. A well-crafted system prompt is crucial for reliable agent behavior.
Model Configuration: LiteLLMModel is a bridge to the language model that acts as the agent's "brain". Strands uses litellm under the hood, which means you can easily switch between dozens of LLM providers (like OpenAI, Anthropic, Google, etc.) just by changing the model_id.
Tool Usage: Tools are the agent's hands and eyes. By giving the agent the http_request tool, we grant it the ability to access the internet. The agent's LLM brain knows how to use this tool to follow the instructions in the system prompt.
Agent Instantiation: The Agent class brings everything together. We provide it with the system prompt, the model, and the tools it's allowed to use.
Invocation: Calling the agent is as simple as calling a function: weather_agent(user_query). The agent takes the query, thinks step-by-step (using the LLM), uses its tools as needed, and returns a final, synthesized answer.

Further Learning

Watch the Video: AWS Strands Course Playlist
Read the Docs: Official Strands Documentation

Previous Lesson	Next Lesson
N/A	Lesson 2: Session Management

README.md

Lesson 1: Your First AI Agent

Key Concepts Explained

Further Learning

Navigation