Daily AI CatchupGames
Reading about a context window and having to actually budget one are different things. Every explainer on this site comes with a small game that makes you make the call yourself — which is the part that sticks.
They take about two minutes each. There is no sign-up, no timer bearing down on you, and no score anyone else can see. Get one wrong and you get told why, which is the whole point.
Games get refreshed as the field moves. A game about model releases written a year ago is not stale, it is wrong — so when the facts change, the game changes with them.
The Prompt on a Budget
Keep only the lines of a prompt that actually change the output. Everything else is just politeness tax.
About 2 min · no sign-up
Fix the Rules
A written instruction is only worth what it can actually enforce. Spot the fixes that close the loophole — not just the ones that sound stricter.
About 2 min · no sign-up
Agent or Not?
Five scenarios. Decide what actually makes each one "agentic" — or not.
About 2 min · no sign-up
How a Skill Gets Used
An agent has dozens of skills loaded and a tiny context window. Put the actual sequence in order.
About 1 min · no sign-up
Pick the Right Model
Sort ten everyday tasks onto the model tier that actually fits them, not the biggest one you can reach for.
About 2 min · no sign-up