Run a Mock Beginner Coding Interview with One Problem and Scored Feedback

Get a step-by-step mock technical interview with one beginner coding problem, a talk-through-your-approach stage, and a scored rubric for clarity, correctness and edge cases.

Prompt

Prompt
You are a friendly but honest software engineer running a 30-minute beginner technical interview.

First, ask me these questions in one message and wait for my answers:
1. Which programming language will I use?
2. How long have I been coding?
3. Any topic to focus on or avoid (strings, arrays, loops, dictionaries)?

Then run the interview one step at a time, waiting for my reply after each step:
Step 1: Give one beginner-level problem with two sample inputs and outputs.
Step 2: Ask me to explain my approach in plain words before coding. Ask at most one follow-up question.
Step 3: Ask me to write the code, then ask which edge cases I would test.
Step 4: Give feedback.

Rules:
- Never reveal the solution or give hints before Step 4 unless I ask; if I ask, give the smallest nudge and mention it in the feedback.
- Score Clarity, Correctness and Edge Cases from 1 to 4, and tie each score to something I actually said or wrote.
- If an input breaks my code, name it exactly.

Output format for Step 4:
- Scorecard table: Criterion | Score (1-4) | Evidence
- What went well (2 bullets)
- Fix next (3 bullets, each a concrete action)
- Clean reference solution with time complexity in one line

End by noting what you could not judge, such as whether my code actually runs.

Example output

At a glanceMock interview scorecard by criterionScores on a 1-4 rubric for the first-unique-character problem
  • Clarity: 3/4
  • Correctness: 3/4
  • Edge Cases: 2/4

Example for: a self-taught learner, 5 months of Python, asked to focus on strings. Problem: return the first character in a string that appears only once, or an empty string if none.

Scorecard
CriterionScore (1-4)Evidence
Clarity3You said "count every letter first, then walk the string again" before coding. Clear plan, but you didn't say what to return when nothing is unique until I asked.
Correctness3Works on both samples ("swiss" → "w", "level" → "v"). Returns None instead of "" for "aabb", which breaks the stated contract.
Edge Cases2You listed an empty string. You missed all-repeating input and mixed case ("Aa"). You asked for one hint about dictionaries; noted here.

Input that breaks your code: "aabb" returns None, expected "".

What went well
  • You chose a two-pass approach on your own and explained why it beats checking every character against every other one.
  • Your variable names (counts, ch) made the code easy to follow.
Fix next
  • Before coding, state the return value for "no answer" out loud and write it as the last line of your function.
  • Practice listing three edge cases every time: empty input, all-same or all-repeating input, and one odd character type (case, spaces, digits).
  • Rewrite this solution using collections.Counter, then explain to yourself why it is still two passes.
Reference solution
def first_unique(s):
    counts = {}
    for ch in s:
        counts[ch] = counts.get(ch, 0) + 1
    for ch in s:
        if counts[ch] == 1:
            return ch
    return ""

Time complexity: O(n) time and O(k) extra space, where k is the number of distinct characters.

What I could not judge: I did not run your code, so I traced it by hand. I also could not judge your pacing, since timing in a chat differs from a live interview.

Written by AI from this prompt. Your results will vary.

Tips and common edits

Why this works

  • Forcing a spoken plan before coding mirrors real interviews, where explaining your approach counts as much as the final code.
  • Pausing after each step keeps the model from dumping the answer and lets you practice under realistic pressure.
  • Tying every score to quoted evidence and a breaking input makes feedback specific instead of generic praise.

Use it in ChatGPT, Claude or Gemini when practicing for a first junior developer or bootcamp interview. Set aside 30 minutes, have a code editor open to test your solution, and answer out loud before typing.

Common edits

  • Ask for two problems back to back and a combined scorecard to simulate a longer screening round.
  • Add a fourth rubric row for time and space complexity once you are past the basics.

Prompt details

By
Prompts-GPT editors
Published

Did this prompt work?

Keep exploring

More in Coding