Build AI Agent Harnesses in a Free Sandbox Workshop
Building AI agent harnesses with the open-source Strands Agents SDK is now something you can practice in a real AWS environment, for free, without setting up an account. This workshop gives you a Builder Center sandbox, walks you through the same patterns Morgan Willis covers in her 14-video YouTube course, and ends with a deployed agent in the cloud.
Building AI agent harnesses with the open-source Strands Agents SDK is now something you can practice in a real AWS environment, for free, without setting up an account. This workshop gives you a Builder Center sandbox, walks you through the same patterns Morgan Willis covers in her 14-video YouTube course , and ends with a deployed agent in the cloud.
No account setup. No cost. Available in English, Spanish, and Portuguese.
Follow the Videos, Then Build
Morgan's course walks through the full arc of building AI agent harnesses with the Strands Agents SDK:
| Phase | Topics |
|---|---|
| Fundamentals | How AI agents work, the agent loop, model providers, tools |
| Intermediate patterns | MCP tools, callbacks, streaming, hooks, plugins, steering, conversation management |
| Advanced | Agents-as-tools, graph workflows, agent swarms, multi-agent orchestration |
| Production | Guardrails, observability, evaluation, cloud deployment |
The workshop picks up where the videos leave off. Instead of watching, you deploy. Instead of following along on a screen, you run real agents in a real AWS environment. Builder Center hands you that environment automatically.
What Is a Builder Center Sandbox?
A Builder Center sandbox gives you a temporary, pre-configured AWS account to run the workshop. No personal account, no cost concerns, and when the session ends it cleans itself up. You show up, open the workshop, and start building.
This makes it useful for team learning sessions, hackathons, and individual self-paced study.
What You'll Build
You'll build an AI agent harness with Strands Agents that:
- Sets up the agent with the Strands SDK and the reasoning loop
- Connects tools, including MCP-based tools
- Uses hooks to control agent behavior
- Implements steering to improve reliability
- Manages context and conversation state
- Deploys to AWS
Each section maps directly to a module in Morgan's course. You can go back to the video for deeper context at any point.
Let's See It in Action
Things to know before you start:
- Language: The workshop is available in English, Spanish, and Portuguese
- Prerequisites: Basic Python knowledge and having watched Morgan's course (or familiarity with Strands Agents)
- Time: Approximately 2 hours
- Cost: Free. The Builder Center sandbox provides the AWS account.
Access the workshop:
- 🇺🇸 English: Open in Builder Center
- 🇪🇸 Español: Abrir en Builder Center
- 🇧🇷 Português: Abrir no Builder Center
Watch the course, then run the workshop. That's the fastest path from curious to building.
📬 Building reliable AI agents? I write about agent memory, guardrails, evaluation, and multi-agent patterns. Subscribe to my newsletter to get the next one.
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article