Keeping your skills

Deliberate practice with the AI switched off, attempting before comparing, using the tool as a tutor rather than an answer machine, and the fundamentals no tool replaces.

Deliberate practice

Also called practice without AI, effortful practice

What: Regular, focused practice slightly beyond what you can already do, with feedback, done without AI completions so the struggle is yours.
Why: The struggle is how skill forms. If the tool always fills the gap, you get the code and skip the learning.
Example: Three evenings a week, solve one problem or build one small feature with AI off, then compare with an AI or reference solution.

AI-off sessions

Also called snooze suggestions, disable completions

What: Blocks of time where suggestions and agents are switched off on purpose, for practice, exams or learning a new language.
Why: Ghost text is hard to ignore. Turning it off is the only reliable way to find out what you can do alone.

Code

# Claude Code only acts when you start it:
# close the session and code in your editor.
# Also turn off any editor completions for the
# practice session.

Attempt first, then compare

What: Writing your own solution, or at least your own plan, before asking the tool, then comparing the two and noting what it did differently and why.
Why: Comparing after an honest attempt teaches far more than reading an answer cold, and shows you exactly where your gaps are.

Watch out

Opening chat "just for a hint" after thirty seconds, before you have really tried.

Tutor mode prompts

Also called Socratic prompting, quiz me

What: Prompts that make the tool ask you questions, give hints and review your reasoning, instead of writing the answer.
Why: The same tool that can make you dependent can also be a patient tutor, if you set the rules.

Code

I am learning, not shipping. Do not write code
for me. Ask me one question at a time about how I
would solve this, give a hint if I am stuck, and
point out mistakes in my reasoning. Only show a
solution if I type "show me".

Read the docs first

Also called official documentation, RTFM

What: Going to the official documentation of a language, library or tool before asking a model about it, especially for anything version-specific.
Why: Docs are current and correct; the model may not be. Reading docs is also a skill in itself, and interviews and jobs expect it.

Fundamentals AI does not replace

What: Data structures and algorithms, how HTTP and databases work, reading unfamiliar code, debugging, Git, testing and writing clear explanations.
Why: They are what you use to judge AI output. Without them you cannot tell good code from plausible code, and coding rounds still test them.

Watch out

Skipping the basics because the tool can write them, then being unable to review the tool.

Explain without notes

Also called Feynman technique, teach it back

What: Explaining a concept or a piece of your own code out loud or in writing, without looking anything up, as if to a classmate.
Why: Gaps show up immediately. It is the fastest test of whether you understood something or only watched it being done.

Go deeper