Article
Diagnose a Broken Game Mechanic Using Logs and Focused Prompts
When a collectible awards several points for one touch, a screenshot of the game rarely explains why. You need the event sequence, script context and the…
Read more →
Article
When a collectible awards several points for one touch, a screenshot of the game rarely explains why. You need the event sequence, script context and the…
Read more →Article
This prototype rewards a player once when they touch an available coin, hides that coin during a cooldown and then makes it available again. The coin…
Read more →Article
“Make a collectible system” can mean several different things. Can every player collect the same coin? Does it respawn? Does the score survive leaving? Agreeing those…
Read more →Article
Three attractive images do not automatically look like a family. Consistency comes from deciding which properties stay fixed and reviewing the results against those decisions.
Read more →Article
A test that passes is reassuring only if it would fail when the behaviour is wrong. AI can generate plausible test code that never checks the…
Read more →Article
An error message is useful evidence. Give an AI assistant the exact failure, the relevant code and the steps that trigger it, and you have a…
Read more →Article
Your first AI-assisted app should be small enough to inspect. A reading list is a useful example because you can observe input handling, changes to the…
Read more →Article
“Build me an app” leaves a lot of decisions to the assistant. It may add accounts, databases and integrations before you have agreed on the basic…
Read more →Article
An AI conversation can give you opportunities to practise, but constant correction can interrupt the conversation and make every answer feel like a test. Set a…
Read more →Article
A lesson plan needs more than an attractive list of activities. The activities must lead toward an observable objective, fit the available time and show whether…
Read more →