CodeStaminaOpen preview

Free · No account · 7–10 minutes

Find the engineering skills AI can hide.

Sixteen questions across eight areas of engineering judgment: reading code, debugging, async reasoning, contracts, review, edge cases, verifying generated code and systems thinking.

WHAT IT MEASURES

Eight areas, scored separately.

A single number hides the thing worth knowing. The result breaks down by area, because the point is to find the one that is behind.

  1. 01

    Code reading

    Tracing what code actually does, rather than what it looks like it does.

  2. 02

    Debugging

    Reasoning from a symptom back to the cause that explains all of it.

  3. 03

    Async reasoning

    Ordering, races and shared state across awaits.

  4. 04

    Types & contracts

    Knowing what a type actually guarantees, and where it stops.

  5. 05

    Code review

    Spotting the behaviour a diff changed but did not mention.

  6. 06

    Edge-case detection

    Finding the input that the happy path never considered.

  7. 07

    AI-output verification

    Checking generated code that reads well and passes its tests.

  8. 08

    Systems thinking

    Reasoning about load, failure and recovery beyond one process.

What this is, and what it is not

Sixteen questions cannot measure an engineer. They can find a direction. Each area gets two questions of different difficulty, and the score is weighted by difficulty rather than by count, so getting the hard one right moves the number more than getting the easy one right.

Treat a low area as a hypothesis worth testing, not a verdict. The questions are drawn from defects that survive review in real codebases — off-by-one reasoning about whitespace, a lost update hiding inside sum += await, a timezone assumption, an assistant’s SQL escaping that is not parameterisation. If one of them catches you, that is the useful outcome.

QUESTIONS

Before you start

Do I need an account?
No. The benchmark runs without signing in, and the result link contains no personal information.
How long does it take?
Sixteen questions, most people finish in seven to ten minutes. There is no timer and no penalty for thinking.
Is this an assessment I can put on a CV?
No, and you should be suspicious of anything that claims otherwise at this length. Sixteen questions find a direction to train, not a certification.
What happens to my answers?
They are scored on the server and discarded. What is stored is the eight scores, how long the run took, and which question set you saw.
Can I see why I got something wrong?
Yes — the full reasoning for every question appears once you finish. It is only shown then, so the answers stay out of the page beforehand.

JOIN THE WAITLIST

Help shape what we build next.

Tell us where you want to start and what you want to practise. We’ll use your answers to understand demand and email you about relevant program availability.

What would you like to practise?

We store your email and answers for this waitlist. No campaign is being sent now. Contact mo@codestamina.com to ask us to remove your entry.