Generates the questions a reviewer asks about your submission, quizzes you one at a time, and grades how well you explain your own tradeoffs.
I submitted a take-home and have a walkthrough call coming up. The assignment:
{{assignment}}
A summary of what I built and the decisions I made:
{{my_solution_summary}}
Run the walkthrough call as the reviewing engineer:
- Ask me to walk through the design in two minutes. Then ask one question per message, at least eight questions, covering: why I chose the data structures I did, what I would change with more time, what breaks at 100x the input, how I tested it, what I would do about a specific edge case you pick from the assignment, and what I would have done differently knowing what I know now.
- If I explain a decision without a reason, ask for the reason. If I say "it was simpler", ask what the alternative would have cost.
- After the questions, grade me on: ownership of decisions, depth on tradeoffs, awareness of limits, and how I handled a question I did not know. Quote my weakest answer and show what a strong version sounds like.
Start with the two-minute walkthrough request.
Have your code open and refer to it by function name; reviewers do. Practice the two-minute walkthrough until it is under two minutes. Note every question you fumbled and prepare a real answer.
Questions that reference specifics from your summary, at least one "why not the alternative" push, and a graded transcript you can rehearse from.
Useful next steps:
