Skip to lesson
Exit
Capstone: Mock Loop & Self-Exam1 / 2

2 min lesson

Present like it's the real room

Make the call in this situation: "On the open prompt 'design Cursor's bug-triage system from scratch,' you realize you don't know how many tickets come in daily. The team is observing. What do you do?" Explain what supports it.

Step 1 of 2

Present like it's the real roomThe last 10 minutes

Close every rehearsal with a crisp recommendation: here's the problem as I scoped it, here's what I built, here's why, here's what I'd do next. Then self-score on the four dimensions this round actually grades.

Say it like this

My ninety-second close on design Cursor's bug-triage system from scratch. The problem as I scoped it: I assumed tickets, Discord and the forum as the inputs, no triage tooling yet, and picked one vertical to prove end to end. What I built: an agent-assisted pipeline that pulls every channel into one queue, clusters and dedups the reports, then scores each cluster by severity × impact into a one-page brief. Why: a human checkpoint after clustering and on the final ranking keeps a quiet-but-catastrophic data-loss bug from getting buried under a loud cosmetic one. What I'd do next: wire ARR exposure into the scoring and ship the first weekly quality brief, then watch whether time-to-resolution actually drops.

Dimension
Autonomy
Pass looks like
Stated assumptions, made calls, moved without permission
Red flag
Asked which approach to take repeatedly
Dimension
Product sense
Pass looks like
Tied the work to a real user-felt quality outcome
Red flag
Built a system with no clear user benefit
Dimension
Depth
Pass looks like
Went one layer past the obvious, with tradeoffs named
Red flag
Stayed surface-level, no hard choices
Dimension
Craft
Pass looks like
Clean, demoable slice you'd be proud to show
Red flag
Sprawling, half-working, presented as a shrug

Score 1–5 on each. The round is graded on your floor, so attack the lowest.