Foundations
What am I actually talking to?
What the model sees, what it charges you, what it can reach on your machine, and the ways a session fails while it still looks productive.
Coding with AI, start to shipped
The whole path in order: what you are talking to, where the work happens, how to direct it, what you are building, how to ship it, and how to keep it running.
What am I actually talking to?
What the model sees, what it charges you, what it can reach on your machine, and the ways a session fails while it still looks productive.
Where does the work actually happen?
Files, the terminal, version control, reading a diff, and running the project on your own machine. The room the work happens in.
How do I turn an idea into shipped changes?
Sizing a piece of work, stating the outcome and the boundary, keeping memory in files, and telling a plausible answer from a proven one.
What are the parts of an app, and which do I need?
The shape of a web app, picking a stack you can finish alone, the interface people actually touch, where data lives, accounts, and other people's services.
What has to be safe before real people arrive?
How secrets leak, the mistakes that expose user data, what you may store about people, and what to never let an AI do unsupervised. Before you deploy, not after.
How does it get on the internet?
Hosting, domains, environments, secrets, and knowing how to undo a deploy before you need to.
It is live and something broke.
Logs, monitoring, backups, the bill, dependency updates, and changing something people are already using.
How do I make the good session repeatable?
Command-line fluency, instruction files, automatic checks, reusable procedures, and splitting one job across several sessions.