Free Lesson

Build an environment aware AI Agent from Scratch

60 min
Jul 20, 2026 5:00 PM
Virtual (Zoom)

In this video

What you'll learn

What's actually inside an agent harness?

You'll see the harness for what it is, a loop simple enough to hold in your head

Can you write an agent's behavior in plain English?

Yes. you'll see how to write skills in plain English so behavior stays transparent, auditable, and correctable

What happens when your agent leaves the terminal?

You'll see the same simple agent wire into any app and carry personal context, the primitive for building real products.

Why this topic matters

Everyone calls Claude Code and Cursor "agent harnesses," a word that makes them sound like complex machinery only a vendor can build. They aren't. The harness is the agent, and it's a loop you can build yourself in an afternoon. Once you see  that, the agent stops being a black box you rent and becomes a primitive you can shape into real products.

You'll learn from

John Berryman

John Berryman

Engineer for GitHub Copilot, author of Prompt Engineering for LLMs (O'Reilly)

Greg Ceccarelli

Greg Ceccarelli

Co-Founder • Ex-CPO Pluralsight • Data at GitHub, Dropbox & Google

Pluralsight
GitHub
Dropbox
Google

Previously at

GitHub
Pluralsight
Dropbox
Google
Dropbox
See all products from Greg