Snag AcademyInspector
It even works… until it doesn't.
How the difficulty is pitched
The fault is latent — it works fine, until the conditions change.
This is real engineering judgement: margins, dependencies, and faults that have been present all along without failing yet.
What a student at this stage can actually do
This band asks for something most school science never does: reasoning about a fault that has been present for months without failing yet. A missing earth wire changes nothing on an ordinary day. An open quick-release survives the driveway. A worn brake pad passes every gentle test.
Engineers call these latent failures, and they are the most dangerous class there is, because the system certifies itself as healthy right up to the accident. Recognising one requires holding two questions apart: does this work, and what happens when conditions change?
The second demand is about margins. A case here is often not about a part at all but about a reserve that has been quietly eaten: attention, trust, tolerance, time.
The four kinds of snag, at this level
- Missing — a safety system that does nothing whatsoever on a normal day, which is exactly why its absence hides for months.
- Wrong — a part that is present, undamaged, and in the wrong state. Gentle testing certifies the lie.
- Extra — something that spends a shared resource nobody is measuring. A false alarm costs no money and spends the credibility of every future alarm.
- Order — a procedure whose waiting step is the real work, and the first thing impatience removes.
Cases at this level
The Password That Worked Fine
Three years, never shared, never guessed. Still wrong.
The Chain That Cries Wolf
Nothing broke. The alarm still works. And every future alarm is now slower.
One Letter Off
A perfect page, a real padlock, and one character in the address.
The Midnight Plan
Nine honest hours a day, and the marks went down.
Open the app — free, no sign-up
Where students usually get stuck
Insisting that some component must be faulty. In several cases the hardware is perfect and the fault is the operating point a person selected. Diagnosing a choice rather than a part is a real conceptual step and it is worth naming out loud when it happens.
Knowing the rule without the mechanism. Most teenagers can recite "check for the padlock" and "do not share an OTP". Far fewer can say what HTTPS actually protects, or why a strong password is irrelevant on an unencrypted connection. These cases are built to find that gap, and they usually do.
Wanting a picture. Several cases here have no drawing at all, deliberately, because the reasoning is not visual. A trust economy has no diagram.
What the first week usually looks like
These are slower. One case can reasonably be a whole sitting, and the Technical Note is the point of the exercise rather than a reward at the end of it. Several notes end with a lesson that generalises well beyond the case, and those are the lines worth arguing about over dinner.
The cases on scams, misinformation and online safety are the ones most worth doing with a parent in the room, not because they are frightening but because the reasoning transfers directly to things arriving on the student's phone this week.
After this band
Inspector is the top band, so the useful next move is not a harder shelf. It is Create, where a student builds their own case: choose the kind of snag, pick a system, hide or break exactly one thing, and write the explanation. Constructing a diagnosable fault is a considerably harder skill than finding one, and it is the clearest evidence that the method has landed.
Alongside the app
- Notes — writing about diagnosis and safety, with every claim checked against its source.
- How to get more than screen time out of it — the "teach it back" habit is worth more than the next case.
Other shelves
👀 Explorer (Class 1–5) · 🔍 Investigator (Class 6–8) · Under 5 years? · High-contrast cards for babies