Strands Agents Series
Overview
This is a series of slide decks I'm putting together to talk about Strands Agents, an open-source SDK for building and running AI agents, at the AWS User Group Wellington. There's an overview deck that tours the whole toolkit at a high level, followed by one deck per topic that goes deep on how it works, how to set it up and how it fits with the rest of Strands.
The decks
| # | Deck | Covers | Hands-on repo |
|---|---|---|---|
| 00 | Overview | What Strands is, the model-driven approach, and a map of the Strands harness, Strands Harness SDK, Strands Shell and Evals SDK | |
| 01 | Getting Started | Python and TypeScript quickstarts and a first agent | |
| 02 | Agent Loop | The agent loop, lifecycle controls, state, prompts and retry strategies | https://github.com/chiwaichan/strands-agents-immersion-day-02-agent-loop |
| 03 | Model Providers | Amazon Bedrock, Amazon Nova, Ollama, LiteLLM and others, model routing and prompt caching | https://github.com/chiwaichan/strands-agents-immersion-day-03-model-providers |
| 04 | Tools | Custom tools, MCP tools and transports, vended and community tools, tool executors | https://github.com/chiwaichan/strands-agents-immersion-day-04-tools |
| 05 | Streaming and Structured Output | Async iterators, callback handlers, events and typed responses | https://github.com/chiwaichan/strands-agents-immersion-day-05-streaming-structured-output |
| 06 | Sessions and Memory | Session persistence, snapshots, storage, long-term memory and the Bedrock Knowledge Base store | https://github.com/chiwaichan/strands-agents-immersion-day-06-sessions-memory |
| 07 | Context Management | Conversation management, built-in modes, custom strategies, context offloader and injector | https://github.com/chiwaichan/strands-agents-immersion-day-07-context-management |
| 08 | Hooks and Plugins | Hooks, hook events, Skills, GoalLoop and custom plugins | https://github.com/chiwaichan/strands-agents-immersion-day-08-hooks-plugins |
| 09 | Interventions | Interrupts, human in the loop, steering and Cedar authorization | https://github.com/chiwaichan/strands-agents-immersion-day-09-interventions |
| 10 | Multi-Agent | Agents as tools, swarm, graph, workflow and agent-to-agent (A2A) | https://github.com/chiwaichan/strands-agents-immersion-day-10-multi-agent |
| 11 | Voice | Bidirectional voice agents, barge-in, I/O, and the Nova Sonic, Gemini Live and OpenAI Realtime models | https://github.com/chiwaichan/strands-agents-immersion-day-11-voice |
| 12 | Harness and Shell | The assembled Strands harness, its built-in tools, sandboxes and Strands Shell | https://github.com/chiwaichan/strands-agents-immersion-day-12-harness-shell |
| 13 | Observability | Metrics, traces, logs and operating agents in production | https://github.com/chiwaichan/strands-agents-immersion-day-13-observability |
| 14 | Evals | Evaluators, simulators, detectors, red teaming, chaos testing and the CLI | https://github.com/chiwaichan/strands-agents-immersion-day-14-evals |
| 15 | Security and Guardrails | Guardrails, PII redaction, trusted message history, prompt engineering and responsible AI | https://github.com/chiwaichan/strands-agents-immersion-day-15-security-guardrails |
| 16 | Deployment | AgentCore, Lambda, Fargate, App Runner, EKS, EC2, Docker, Kubernetes and Terraform | https://github.com/chiwaichan/strands-agents-immersion-day-16-deployment |