🎯 Career

IT Interview Prep Coach prompts

Five prompts to walk into an IT interview ready — mock technical drills, STAR behavioral answers, saying cert concepts out loud, salary negotiation, and a structured post-interview debrief.

Tested 2026-05 Claude 4.7 OpusGPT-5Gemini 2.5 Pro #interview#career#cert#jobs
Honest note — A model is a sparring partner, not the interviewer. It will be more patient and more predictable than a real panel. Use these to build reflexes, then expect the real thing to be messier.

Prompts in this set

  1. 1. Run a realistic mock technical interview
  2. 2. Turn raw experience into STAR behavioral answers
  3. 3. Drill explaining a cert concept out loud
  4. 4. Build a salary negotiation script
  5. 5. Post-interview debrief + follow-up email

1. Run a realistic mock technical interview

When you can recall facts but freeze the moment someone asks you a question live.

Claude 4.7 Opus (2026-05)GPT-5 (2026-05)
Act as a technical interviewer for a <ROLE> position (e.g. Junior Cloud Engineer, Network Technician, SOC Analyst).

My background: <CERTS_AND_EXPERIENCE>
Target seniority: <JUNIOR | CONFIRMED>

Run a mock interview with these rules:
1. Ask ONE question at a time. Wait for my answer before continuing.
2. Mix question types: 2 fundamentals, 2 scenario/troubleshooting, 1 "explain this to a non-technical colleague".
3. After each of my answers, give: a score /5, what was missing, and the answer a strong candidate would give — in under 120 words.
4. Escalate difficulty only if I score 4+.
5. After 5 questions, give a final readiness verdict and the single weakest area to drill.

Start with question 1 now. Do not show me all the questions up front.
TipAnswer out loud or in full sentences — typing two words defeats the purpose. If it goes too easy, tell it to assume a stricter panel.

2. Turn raw experience into STAR behavioral answers

When you have done the work but can't tell the story — behavioral rounds reward structure, not modesty.

Claude 4.7 Opus (2026-05)GPT-5 (2026-05)
Help me build a STAR (Situation, Task, Action, Result) answer for a behavioral interview question.

Question I expect: <BEHAVIORAL_QUESTION> (e.g. "Tell me about a difficult problem you solved")
Raw material — what actually happened: <DUMP_EVERYTHING_YOU_REMEMBER>
Role I'm interviewing for: <ROLE>

Do this:
1. Reshape my raw material into a tight STAR answer, ~150 words spoken aloud (~60 seconds).
2. Make the Result concrete — quantify it if my material allows, never invent numbers.
3. Flag any part that sounds like a team accomplishment I'm claiming solo.
4. List 2 likely follow-up questions and a one-line answer for each.

If my raw material is too thin to support the question, say so and tell me what detail to recall.
TipBe honest in the raw dump — if the model has to invent a result, the story collapses under the first follow-up question.

3. Drill explaining a cert concept out loud

Interviewers love "explain <concept> to me" — knowing it on a multiple-choice test is not the same as saying it cleanly.

Claude 4.7 Opus (2026-05)GPT-5 (2026-05)
I'm going to explain a technical concept out loud as if in an interview. Grade me like a hiring manager, not a teacher.

Concept: <CONCEPT> (e.g. "the difference between a subnet and a VLAN", "how TLS works", "what IAM roles are")
Role context: <ROLE>

Here is my explanation:
<PASTE_YOUR_EXPLANATION>

Grade it on:
1. Correctness — any factual error, however small.
2. Clarity — would a slightly-less-technical interviewer follow it?
3. Structure — did I lead with the point or bury it?
4. Length — interview answers should be 30-60 seconds; flag rambling.

Give a score /5, then a model answer of the same length so I can hear the gap. End with one follow-up question an interviewer would ask to test if I really understand it.
TipRun this 3 times for the same concept with different phrasings — the goal is a fluent answer, not a memorised script.

4. Build a salary negotiation script

When the offer arrives and "uh, sure, that works" is about to cost you thousands a year.

Claude 4.7 Opus (2026-05)GPT-5 (2026-05)
Help me prepare a salary negotiation for an IT job offer.

Role + location: <ROLE_AND_CITY_COUNTRY>
My situation: <CERTS, YEARS_OF_EXPERIENCE, CURRENT_SALARY_OR_NONE>
The offer on the table: <OFFERED_SALARY_AND_ANY_BENEFITS>
Market reference I have: <SALARY_RANGE_YOU_RESEARCHED_OR_"NONE">

Produce:
1. A realistic target number and a walk-away floor, with one line of reasoning each.
2. A 4-sentence negotiation script — appreciative, anchored on value, specific number, open question.
3. Two likely pushbacks from the employer and a calm reply to each.
4. Non-salary levers to ask for if the number truly can't move (remote days, training budget, exam fees, review date).

Keep the tone collaborative, not combative. If my offer already looks strong for the market, tell me honestly.
TipFeed it a real market range — from a salary page or a recruiter — or it will hedge. Never quote a number you haven't sanity-checked.

5. Post-interview debrief + follow-up email

The 30 minutes after an interview is when the lessons are still sharp — capture them before they fade.

Claude 4.7 Opus (2026-05)GPT-5 (2026-05)
I just finished an interview. Help me debrief while it's fresh.

Role: <ROLE>
What went well: <NOTES>
Where I struggled or blanked: <NOTES>
Questions they asked that surprised me: <NOTES>

Do three things:
1. Turn my struggles into a concrete drill list for the next interview — be specific ("revise OSPF areas", not "study networking").
2. Draft a short, genuine thank-you / follow-up email — 5 sentences max, references one real moment from the conversation, no clichés.
3. Give an honest read on how it likely went and what signal to watch for next.

Don't be a cheerleader. If my notes suggest it went badly, say so and focus on the next one.
TipRun this the same day. Paste the drill list into your study planner so the next prep round starts from real gaps.

How to use these prompts

Each prompt has placeholders in <ANGLE_BRACKETS> — fill them in before pasting. Copy the prompt with the button, paste into Claude, ChatGPT, Gemini, or any chat-UI'd LLM.

Why "model tested" dates matter

LLMs improve and regress with every release. A prompt that worked on Claude 3.5 may need rewriting for Claude 4. The dates show when each prompt was last verified — anything older than 6 months should be re-tested before depending on it.

Found a better prompt?

Hit contact and share — we keep prompts that beat ours.