Claude Prompts for Hiring Decisions That Hold Up
20 copy-paste Claude prompts for hiring: interview design, structured evaluation, debrief facilitation, calibration prep, reference checks, and offer construction. Long-context Claude shines on multi-interview synthesis.
In short: This page contains 16 copy-paste ready prompts, organized into 4 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.
Interview Design
4 promptsInterview Loop Design
1/16✨ What it does
Claude designs an interview loop for [role + level] with stages, competencies per stage, format, time, who interviews, and what each stage filters, without redundant rounds. Fill your role and level, then cut any stage that tests the same thing twice.
Design interview loop for [role + level]. Output: stages (recruiter screen, hiring manager, peer panel, leadership, exec final), competencies tested per stage, interview format per stage (behavioral, technical, case, work sample), time per stage, who interviews, what each stage filters for. No redundancy.
Pro tip: Interview loops with 3 panels asking same behavioral questions = exhausted candidate + no new signal. Each stage tests distinct competency = efficient + thorough.
Behavioral Question Generator
2/16✨ What it does
Claude writes 5 STAR behavioral questions for [competency], plus what a good or concerning answer shows, two follow-up probes each, and alternate phrasings if the candidate stalls. Name your competency, then use the probes instead of moving on after the first story.
Generate behavioral interview questions for [competency]. Output: 5 STAR-format questions, what good answer reveals, what concerning answer reveals, 2 follow-up probes per question to verify depth, alternative phrasings if candidate doesn't engage. Specific to competency, not generic.
Pro tip: Specific behavioral questions = specific signal. "Tell me about leadership" = vague. "Tell me about a time you led without authority + how you got buy-in" = competency-specific = useful answer.
Work Sample Design
3/16✨ What it does
Claude designs a work sample for [role] under 4 hours that simulates the real job, with a brief, materials, time guidance, deliverables, and a rubric. Name the role, then time the assignment yourself before you send it to candidates.
Design work sample for [role]. Constraints: under 4 hours candidate time, simulates real job, evaluatable against rubric. Output: brief, materials provided, time guidance, deliverables expected, rubric for evaluation. Test the actual job, not unrelated puzzles.
Pro tip: 4 hours = ethical max. Real-job simulation > leetcode trick. Rubric pre-built = consistent evaluation. Common mistake: design assignment, evaluate gut-feel = bias.
Technical Interview Question Calibration
4/16✨ What it does
Claude calibrates [Paste technical interview questions] for [level]: difficulty, consistency across candidates, leak risk, job relevance versus trivia, the signal of each question, and recommended swaps. Paste your questions, then swap any that are trivia or already leaked.
[Paste technical interview questions]. Calibrate difficulty: appropriate for [level], same difficulty across candidates? Are questions known via leak? Are they real-job-relevant or trivia? What signal does each question give vs what we need to know? Recommend swaps.
Pro tip: Same questions for 6 months = leaked. Refresh quarterly. Trivia questions test memorization, not capability. Real problems > textbook problems for actual signal.
XML tags are just the start. Learn the full Claude workflow.
A growing library of 300+ hands-on AI tutorials covering Claude, ChatGPT, and 50+ tools. New tutorials added every week.
Candidate Evaluation
4 promptsInterview Notes to Scorecard
5/16✨ What it does
Claude turns [Paste interview notes] into a scorecard with a 1-4 score and anchor per competency, quoted evidence, remaining gaps, and a lean recommendation, meant to be scored before debrief. Paste the notes, then lock your scores before the room discusses.
[Paste interview notes]. Convert to structured scorecard. Output: per competency tested, score 1-4 with anchor description, evidence from interview supporting score (specific quotes/moments), gaps still to assess, overall recommendation lean. Score before debrief, not during — order matters.
Pro tip: Scoring during debrief = group bias. Individual scoring before debrief = independent signal. Compare scores in debrief = real calibration. Order is the bias guardrail.
Resume + Background Pre-Screen
6/16✨ What it does
Claude pre-screens [Paste resume] for [role] with a 1-5 experience score, gaps to probe, claims to verify, the interview format that would signal best, and what is unusual. Paste the resume, then verify the title and outcome claims you care about.
[Paste resume]. Pre-screen for [role]. Output: relevant experience score 1-5, gaps to probe in interview, claims to verify (specific titles, scope, outcomes), interview format that'd signal best, what about this candidate is unusual (good or concerning).
Pro tip: Pre-screen by gut = bias. Pre-screen by competency rubric = signal. Specific gaps to probe = focused interview, not exploration. Time-saver + bias-reducer.
Reference Check Synthesis
7/16✨ What it does
Claude synthesizes [Paste reference call notes from 2-3 references] into confirmed strengths, shared concerns, one-off comments that may be bias, surprises, and alignment with interview signals. Paste all notes, then treat a concern from two references as the one you discuss.
[Paste reference call notes from 2-3 references]. Synthesize patterns: confirmed strengths (multiple references), patterns of concern (multiple references), one-off feedback (single reference, possible bias), what surprises me, what aligns with interview signals.
Pro tip: Single reference = noise. Multiple references = signal. Patterns across references = deepest insight. Generic "great employee" answers from all = lazy referencing; specific consistent feedback = real signal.
Take-Home Evaluation
8/16✨ What it does
Claude evaluates [Paste take-home submission] on correctness, quality, judgment, how they explained the approach, what they prioritized and cut, and time management, with a 1-4 per dimension and follow-up topics. Paste the work, then bring your follow-up questions to the next call.
[Paste take-home submission]. Evaluate against rubric: technical correctness, code/work quality, judgment shown, communication of approach, what they prioritized + cut, time management. Score 1-4 per dimension. What I'd want to discuss in follow-up.
Pro tip: Take-home eval = code quality + judgment shown. Working solution with poor code = junior thinking. Elegant solution to wrong problem = direction problems. Both matter.
Debriefs + Calibration
4 promptsDebrief Facilitation Script
9/16✨ What it does
Claude writes a debrief script for [candidate] with a 15-plus-30 agenda, lowest score speaking first, concerns to dig into, a decision protocol, and who breaks a tie. Name your candidate, then start with the lowest score so concerns are not buried.
I'm facilitating debrief for [candidate]. 4 interviewers, scores submitted independently. Output: agenda (15 min individual + 30 min discussion), order of speaking (lowest score first to surface concerns), specific concerns to dig into, decision call protocol, who has tiebreaker.
Pro tip: Lowest score first = concerns surfaced early; not buried by groupthink. Highest score first = others suppress doubts. Order matters more than people realize.
Calibration Across Candidates
10/16✨ What it does
Claude calibrates [Paste 3-5 candidate scorecards] for consistency, interviewer leniency, where scores diverge and why, and who is strongest overall and on which dimensions. Paste the cards, then settle the calibration before you make the offer call.
[Paste 3-5 candidate scorecards]. Calibrate: are scores consistent across candidates? Are interviewers stricter/lenient consistently? Where do scores diverge significantly + why? Which candidate is strongest holistically + on what dimensions? Help me calibrate before final decision.
Pro tip: Same interviewer scoring 3 candidates 4-4-4 = no calibration. 4-3-2 = signal. Knowing rater patterns + adjusting = better hiring than naive averaging.
Tie-Breaker Decision
11/16✨ What it does
Claude forces a choice between two qualified people using [Paste both scorecards]: what each does better, weaknesses, fit for the role today versus later, team chemistry, and which is the safer versus riskier call. Paste both of your cards, then pick one name; do not keep both in limbo.
2 candidates remain, both qualified. [Paste both scorecards]. Help me decide: what each does better than the other, where each is weaker, which fits role today + which fits role tomorrow, which fits team chemistry, which is risk-taking call vs safe call. Force a choice.
Pro tip: Two-good-candidates problem = force the choice on specific dimension that matters most. "Both great" = no signal. "X is better at Y because" = decision. Neither is wrong; one is more right for this hire.
No-Hire Decision Articulation
12/16✨ What it does
Claude helps you articulate a no-hire for [candidate]: the skill, experience, or fit gap, what you needed and did not see, what would change your mind, and feedback you could share, in a form you can defend internally. Name the candidate, then write the gap in one sentence before you send the rejection.
I'm saying no to [candidate]. Help me articulate why specifically: what gap (skill / experience / fit), what alternative I needed but didn't see, what would change my decision (if anything), what feedback I could share with candidate. Decision should be defensible internally.
Pro tip: Vague no-hire ("not the right fit") = bias risk + unhelpful for candidate. Specific gap articulated = bias-checked + credible internally + defensible if challenged.
These prompts give you the what. Tutorials give you the why.
Learn when to use extended thinking, how to build Claude Projects, and workflows that compound. 300+ tutorials and growing.
Offer + Closing
4 promptsCompensation Offer Build
13/16✨ What it does
Claude builds an offer for [candidate] from current pay, market data, internal equity, and level, with base, equity, bonus, signing bonus, benefits in dollars, total-comp transparency, flexibility, and opening versus final room. Fill the candidate, then decide your opening number before you get on the call.
Build offer for [candidate]. Inputs: their current comp, market data, internal equity, level. Output: base, equity, bonus, signing bonus if any, benefits worth $, total comp transparency, where I have flexibility, opening offer vs final-room.
Pro tip: Opening offer = give yourself room. Tight offer = no negotiation room = candidate either accepts (good for us, bad signal for them) or pushes back (we have nothing left). Build slack.
Offer Letter + Pitch
14/16✨ What it does
Claude writes an offer letter and closing pitch for [candidate] with a personalized opener, a one-paragraph role pitch, comp summary, signing logistics, and a response timeline. Name your candidate, then send it warm and clear; they are not hired until they sign.
Offer letter + closing pitch for [candidate]. Personalized opener referencing what excited them, role pitch (one paragraph, why this role specifically), comp summary, signing logistics, expected response timeline. Warm but professional. Final piece of selling — they're not in yet.
Pro tip: Offer = legal doc + sales pitch. Pure legal = cold; pure sales = unprofessional. Both = candidate receives "we want you" in tone they want to feel. Most offer letters fail at the pitch.
Counter-Offer Response
15/16✨ What it does
Claude drafts your response when a candidate gets a counter from their current employer: validate the choice, why your offer and opportunity still fit, what the counter is really about, and permission to decide what is right. Use the draft on the call, then stop pushing if they choose to stay.
Candidate received counter-offer from current employer. Help me respond: validate the choice they're facing, articulate why our offer + opportunity is the right move, address what counter-offer is actually about (money, fear, sunk cost), give them permission to decide what's right.
Pro tip: Counter-offer = current employer realizing they should've treated them better all along. 70% who accept counter-offer leave within a year anyway. Help candidate see the pattern.
Hiring Process Post-Mortem
16/16✨ What it does
Claude runs a hiring post-mortem for [role] on what worked, what slowed you, candidate friction, where a competitor would have been faster, and what to change next time. After the hire lands, pick one process change and apply it on the next req.
Hire just landed for [role]. Help me run post-mortem: what worked in our process, what slowed us down, where candidate feedback flagged friction, where competitor would have moved faster, what to change next time. Process improvement compound across hires.
Pro tip: Hires-as-data = process improves. Each hire teaches: stage too long, question too vague, debrief inconsistent. Process compounds; first hire = learning curve, tenth hire = polished.
Free tool
Prompt Optimizer
Turn a rough idea into a structured, professional AI prompt.
Frequently Asked Questions
Prompts are the starting line. Tutorials are the finish.
A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.
7-day free trial. Cancel anytime.