What this prompt does
This prompt runs a realistic 45-minute mock coding interview where the model acts as the interviewer the entire time and won't solve the problem for you unless you ask. It follows a six-step protocol: present one medium problem in a chosen pattern, expect you to think out loud and nudge you if you go quiet, give a capped number of hints, push for time and space complexity before accepting a solution, ask an extending follow-up, and grade you on communication, correctness, efficiency, and edge cases.
Four variables tune the session. [pattern] selects the problem type, like sliding window. [level] sizes the difficulty to your target seniority. [language] is the language your solution is evaluated in. [hint_tolerance] caps how many hints you can receive, escalating only when you're genuinely stuck. The signal the interviewer watches for is whether you narrate your thinking and handle edge cases, not whether you nail it instantly — and the complexity step is where real interviews separate candidates. Capping hints matters too, since leaning on them early is exactly the crutch a real loop won't give you and the scorecard will quietly reflect.
When to use it
- You want timed, realistic interview reps on a specific
[pattern] - You tend to go silent while coding and need to practice narrating
- You want hints capped so you don't lean on them too early
- You want to be pushed on time and space complexity every time
- You need an extending follow-up to practice optimizing under pressure
- You want a scored breakdown across communication, correctness, efficiency, and edge cases
Example output
The session runs turn by turn: a single medium problem from your chosen pattern sized for your level, prompts to keep you thinking aloud, hints capped at your tolerance, a push for complexity analysis before your solution is accepted, and one follow-up that extends or optimizes the problem. At the end you get a scorecard rating you 1-5 on communication, correctness, efficiency, and edge cases, each with one concrete improvement.
Pro tips
- Ask it to be strict on the complexity step every time; pushing for time and space complexity in
[language]is where real interviews separate candidates - Set
[hint_tolerance]low so you practice working through stuck moments instead of leaning on hints early - Think out loud constantly; the scorecard rewards narrating your reasoning, not silent correctness
- Match
[pattern]and[level]to what you're actually weak on so the reps target real gaps - Treat edge cases as part of the answer, not an afterthought; they're a graded dimension
- Use the extending follow-up to practice optimizing, since interviews often pivot to a harder variant
