learn
Learn the platforms that change your work.
You do not have to be a developer, but the technical step is real: you learn to steer the major AI platforms, with your own project as practice material.
Want your project live first? Start with the guide from prototype to live, or with the guide for your build tool: Lovable, Bolt, v0 or Replit. Background per topic, with sources and dates, is in the knowledge base.
three tracks that run into each other
Learning paths with the first three steps as a preview.
1 · Mastering the AI tools themselves
How Claude Code works: setting up a project, thinking in tasks, giving good instructions, avoiding pitfalls. How Codex and similar agent tools work. And when to reach for Claude, Gemini or the OpenAI models. Practical and concrete, with real examples.
2 · Learning to prompt well
No magic spells, but the way of thinking: breaking a problem down, giving context, asking for precision, checking results and iterating. The skill that makes the difference between mess and progress.
3 · From prototype to product
Putting the tools from the first two tracks to work within your own basestep journey. The bridge between learning and doing: what you learn in the lesson, you apply straight away, with the libraries and templates alongside.
how a learning path works
An ordered series of skills, each with a ready-to-paste AI prompt.
A learning path strings skills together in the right order.An ordered series, not a loose collection.
Every skill comes with a ready-to-paste AI prompt.You paste it into your own AI tool.
The first three steps are a preview.After that, the learning paths are part of Premium.
Every skill carries a review interval.The editors keep the content up to date.
the anatomy of a skill
Every skill has the same structure.
A skill is a short hands-on piece: the practical explanation, the ready-to-paste AI prompt for your own AI tool, and the review interval the editors use to keep the content up to date.
the practical explanation
the ready-to-paste AI prompt
the review interval
A short hands-on piece: practical and concrete, with real examples.
“Paste this prompt into your own AI tool…”
The editors keep the content up to date.
the knowledge base
Type what you are looking for, the knowledge base filters as you go.
The knowledge base is filled from a fixed list of sources that brings in news daily, but nothing goes in unless a person has edited and published it. In the platform you open a search window over it with a keyboard shortcut. Below is the same principle with example items, so you can see how searching feels before you create anything.
knowledge baseexample items, not real content
6 example items
Connecting a domain without panic
Contrast that also works on a phone
Keys that do not belong in your repository
Storing passwords without inventing how yourself
An empty state that does not read as an error
No example item with that word. In the platform, the same window also searches your skills, your UI components and your templates.
What the search window does not do yet: your learning paths are not in it. You do find individual skills, UI components, templates, knowledge items and your own board tasks. Projects, clients, invoices, uploads, Signal and the digest are not in it either. A real command palette with actions in it does not exist yet.
e-learning v3in development
Once it exists, you will read about it here.
E-learning with video, tests and certificates is in development. Today there is no video, no test, no certificate and no progress tracking.
certificate of completion
E-learning will come with a certificate of completion. This is a proof of participation from basestep. It is not a recognised training programme, not a diploma and not an accredited certificate.
The journey starts with a prompt.
The first three steps of a path are a preview; after that, the learning paths are part of Premium.