Frameworks 🧩

We explore ways of working with AI in real day-to-day work.

Then turn what we learn into simple, usable frameworks to help you move beyond just prompting.

Most frameworks are designed to be used with our toolkits, which include checklists, scripts, and templates, that you can use in your own projects.

The four modes of working with agents

Most people end up using AI in just two ways: quick tasks or vague exploring. The Four Modes Framework gives you a clearer map – Transaction, Curious, Builder, and Collaboration – so you can see where you are now, avoid Panic Mode, and design simple ladders towards working with an agent, not just prompting it.

Four Modes Framework →

Take the Modes Quiz

Rails, boundaries & safety

To work with an agent safely and confidently, you need some simple, explicit rules called rails. This framework gives you three types – Scope, Boundary, and Process. Together they clarify what AI is for, what always stays human, and how you’ll work together day to day. It’s a practical starter for collaboration – and for resetting when things get stressful.

Rails, Boundaries & Safety Framework