AI AI Tools Guidebook
Prompt Library

Interview Debrief Prompts: Pull a Drill Plan From Every Round

11 copy-ready prompts for an honest interview debrief — same-day raw recall, question-by-question scoring, cross-loop pattern detection, and a drill plan tied to your real weak spots.

Published: Updated: Author: AI Productivity Guide Team 🌐 查看中文版本

Most candidates skip the debrief, so a five-round loop teaches them roughly the same lesson five times. The honest debrief is hardest right after a bad round: your brain is sorting between “I was nervous” and “I genuinely lack the skill,” and without a structured pass you guess wrong and drill the wrong thing. These 11 prompts force a same-day raw recall, then a question-by-question replay, then cross-interview pattern detection, so the next round is measurably stronger. Pair them with the mock interview prompts to drill the gaps you surface here.

TL;DR

  • Run prompt 1 the same day, within an hour, before memory decays — that is the single highest-leverage habit.
  • After 4+ rounds, run prompt 5 to find your real failure axis (technical depth, communication, scope, or role-fit), then drill only that.
  • Replace every [bracketed] placeholder with your own text. Paste your raw recall; do not pre-polish it.
  • Use a tool with persistent memory so the AI sees the whole loop, not one isolated chat (Claude Projects or a ChatGPT custom GPT — see below).

Which AI to run these in (as of June 2026)

These prompts work in any current chat model, but two setups matter:

  • Claude Pro ($20/mo, Sonnet 4.6) with a Project — create one Project per company loop, drop each debrief in it, and Claude reads across rounds for free. Best for the cross-interview pattern prompts (5, 7) because it keeps the whole loop in 1M-token context.
  • ChatGPT Plus ($20/mo, GPT-5.5) with Memory on — fast turnaround for same-day debriefs; turn on Memory so it recalls your target role and level across chats.

One privacy note: a debrief contains the company name, your weak answers, and recruiter feedback. In ChatGPT, Settings → Data Controls → turn off “Improve the model for everyone” so your transcripts are not used for training; Claude does not train on your chats unless you opt in. Strip any other person’s name before pasting recruiter feedback.

For the practice loop that turns these findings into reps, see mock interview prompts and behavioral question prompts.

Best for

  • Same-day post-interview journaling
  • Multi-round prep loops at a single company
  • After-rejection learning before the next application
  • Promotion / level-up panel debriefs
  • Spotting patterns across 4+ interviews in a month

1. Same-day raw debrief

I just finished a [round, e.g. onsite system design] interview at [company]. Help me debrief in 200 words. Format: (a) the 1 strongest answer + why, (b) the 1 weakest answer + what I lacked, (c) the 1 question I did not expect, (d) the next-round prep I should prioritize.

2. Question-by-question replay

Below is my recall of the interview questions and my answers. For each, rate my answer on a 1-5 scale, name the single missing element, and suggest a 50-word improved version.

[paste questions + answers]

3. After-rejection forensic

I got rejected from [company] after the [round]. Recruiter feedback: "[paste verbatim]". Help me extract: (a) the actual signal they may have seen, (b) which of my interview patterns likely caused it, (c) the 2 specific habits to drill before the next loop.

4. STAR completeness audit

Below are 3 STAR answers I gave today. For each, identify which letter (Situation, Task, Action, Result) was weakest, and rewrite the answer to balance it.

[paste 3 answers]

5. Cross-interview pattern detection

I have done 5 interviews in the past month with these results: [paste outcome per company]. Help me detect the pattern: am I losing on technical depth, communication, scope of ownership, role-fit, or company-specific signal? Be specific and cite which rounds point to each.

6. “What would a senior version of me have said” pass

Below is my answer to question "[Q]". Rewrite it as if I were 1 level more senior than I am today. Mark which words/phrases signal seniority. Then mark which I could realistically deliver in a future interview.

[paste answer]

7. Behavior-vs-technical balance check

Across my last 4 interviews, the scores I have heard: behavioral [x], technical [y], role-fit [z]. Help me decide what to drill next — more technical reps, behavioral storytelling, or company-research depth. Justify with the pattern, not a generic answer.

8. Reverse-question quality

Below are the 3 questions I asked the interviewer. Evaluate each: did it signal seniority, curiosity, role-fit, or wasted time? Suggest 3 stronger questions for the next round of this [role] at [company].

[paste 3 questions]

9. “I bombed it” recovery plan

I clearly bombed today's [round]. Help me write a 100-word self-talk debrief that is honest but does not spiral. End with the 1 specific drill to do tonight before tomorrow's next round.

10. Hiring-manager match analysis

Below is what I know about my hiring manager: [paste bio / LinkedIn / conversation notes]. After meeting them, here is what I noticed: [paste notes]. Help me identify what they value, what concerns I might trigger, and how to address those concerns in the final round.

11. Promotion-interview debrief

I just did a promotion panel for [current level → target level, e.g. senior → staff]. Help me debrief: did I show staff-level scope, cross-team influence, written artifacts, and strategic thinking? Score each on 1-5 with what was missing.

How to chain them in one loop

  1. Within an hour: prompt 1 (raw recall) while details are sharp.
  2. Same evening: prompt 2 or 4 (question replay or STAR audit) to convert recall into specific fixes.
  3. If rejected: prompt 3 (forensic) on the recruiter feedback.
  4. After 4+ rounds: prompt 5, then 7, to find the one axis you keep losing on.
  5. Before the next round: turn the top weakness into a drill with the mock interview prompts.

A debrief without a drill plan is just journaling. Every run should end with one named thing to practice before the next interview.

Common mistakes

  • Skipping the debrief, then repeating the same hesitation in the next round
  • Only debriefing what went well — the failure modes are where the lift lives
  • No drill plan tied to the weakness, so the next interview is the same gamble
  • Treating one bad round as a referendum on your career instead of a single data point
  • No pattern detection across 4+ interviews, so you can’t tell nerves from a real gap
  • Confusing “I was nervous” with “I genuinely don’t know this” and drilling the wrong one
  • Pasting recruiter feedback or a panelist’s name into a chat that trains on your data

FAQ

When should I run the first debrief prompt? Within an hour of leaving the room. Recall decays fast, and the specific phrase you fumbled is the thing worth fixing. Prompt 1 is built for raw, unpolished recall — do not wait until you “feel ready,” because by then you have already rationalized the round.

Should I give the AI the job description and my resume too? Yes for the pattern prompts (5, 7) and the hiring-manager prompt (10). Drop them into a Claude Project or attach them in ChatGPT once, and the model can tell role-fit gaps apart from skill gaps. For the same-day prompt (1), keep it lean so the model focuses on what just happened.

Is it safe to paste recruiter feedback and the company name? Treat it as sensitive. In ChatGPT, turn off model-training under Settings → Data Controls; Claude does not train on chats by default. Remove any individual panelist or recruiter name before pasting — you want the signal, not their identity in a transcript.

Which model handles the cross-interview pattern detection best? Anything with large context and memory. Claude Sonnet 4.6 and GPT-5.5 both hold 1M tokens, so a full loop of five debriefs fits in one conversation. The deciding factor is keeping all rounds in one place — a Claude Project or a single long ChatGPT thread — so prompt 5 has the whole picture.

What do I do with the output? End every debrief with exactly one drill. Feed that weakness into the mock interview prompts and run two or three reps before the next round. The debrief only pays off if it changes what you practice tonight.

Tags: #Prompt #Job search #Interview

Related Articles