Your builder path
0 / 6 chapters- 01
Get your tools ready.
Everything you need before your first project.
- 02
Make the starter yours.
One download becomes the home of your build.
- 03
How your coach plans the build.
Read this while the coach asks you questions.
- 04
How your coach builds it.
Read this while the cards get built.
- 05
Prove that it works.
A working result is what earns the finish.
- 06
Show what you made.
Finish with a result you can explain.
Reference
What KDBM Lite is
KDBM Lite is the KrackedDevs Build Method, cut down to what a beginner needs in one session. Six chapters take you from an empty folder to a build you can demonstrate and explain.
It works in VS Code, Antigravity, Cursor, or any coding workspace whose AI agent can read and write project files and run terminal commands.
Core rule
AI gives you a draft. You turn it into proof.
Everything in this path exists to close the gap between something that looks finished and something you have actually tested.
The loop underneath
Every chapter runs the same four beats at a different scale.
| Step | What you do | What it produces |
|---|---|---|
| Spec | Say what good means | The brief, design, and blueprint |
| Build | Implement one Work Card | One working slice |
| Check | Test the main action | Evidence it works |
| Ship | Save the proof | Something you can show |
What you leave with
- A project folder holding your brief, design, blueprint and build status.
- One working build with a main action you tested yourself.
- Proof you can show: a live link, a recording, or a screenshot.
- An explanation of what you decided, what you checked, and what is next.
Session shape
A live class runs about three hours. Working alone, expect a similar span with breaks where you need them.
| Time | Outcome |
|---|---|
| First 30 minutes | Tools ready and the starter open |
| Next 45 minutes | A plan and a first Work Card you approved |
| Next hour | The main action built, card by card |
| Last 30 minutes | Checked, published or captured, and explained |
Before you paste anything
Use safe, sanitized examples. Never paste passwords, tokens, API keys, client data, or identity numbers into an AI tool.
Common mistakes
- Treating the first AI output as a finished product.
- Choosing a project so broad that nothing can be finished today.
- Building three Work Cards before checking the first one.
- Keeping decisions in chat instead of in the project files.