Most candidates apply to everything or nothing. Neither works in 2026: a detectable applicant tracking system sits in front of roughly 97.8% of Fortune 500 openings (Jobscan’s 2025 ATS report), about 75% of resumes are filtered before a human sees them, and the applicant-to-interview rate fell to around 3% in 2024 from 8.4% a year earlier as AI-assisted volume flooded every posting. Spraying applications is now noise. The fix is triage: read each JD honestly before you spend an hour tailoring, and only invest where you actually fit.
The 12 prompts below do that triage. They separate real requirements from the wishlist, surface hidden signals, and end with a defensible “apply / pivot / pass” call so you stop wasting effort on roles you were never going to clear.
TL;DR
- Use these prompts to decide whether to apply at all, before you tailor your resume or write a cover letter.
- The decision spine is three templates: #1 real vs wishlist, #2 fit score (1-5 across 7 dimensions), #6 apply / pivot / pass.
- Paste your actual background. Asking AI to “be honest” with no record to work from just produces confident guesses.
- For the analysis itself, Claude Opus 4.7 or Sonnet 4.6 tends to hedge instead of inventing numbers; GPT-5.5 is handy when you only have a screenshot of the JD. Verify every figure either model hands you.
- Tailoring where you genuinely fit lifts response rates several-fold over generic blasts — the triage is what makes tailoring worth doing.
Who this is for
Job seekers triaging a long shortlist, career switchers calibrating how far to stretch, and recruiters helping candidates self-select out of obvious mismatches.
Which model to run these in (as of June 2026)
You do not need a paid plan for triage; the free tiers handle a single JD fine. Where the models differ is how they behave when they lack information:
| Model | Plan to access | Best for here | Watch for |
|---|---|---|---|
| Claude Sonnet 4.6 / Opus 4.7 | Free (Sonnet, limited) / Pro $20 / Max $100+ | Honest fit analysis, gap-finding, long JDs | None major; it tends to flag missing data |
| GPT-5.5 | Free / Go $8 / Plus $20 | Pasting a JD screenshot, brainstorming angles | Will invent a comp figure or date if you let it |
| Gemini 3.1 Pro | Google AI Pro $19.99 | Cross-checking a second opinion, Workspace users | Same: confirm any number against a source |
In May 2026 head-to-head testing, when asked to strengthen a claim without a real number, Claude inserted a bracketed placeholder and asked for the figure about 80% of the time, while ChatGPT supplied a plausible invented number roughly 60% of the time with no flag. For fit analysis, that hedging is a feature. Pick whichever model you already pay for and apply the same rule to all of them: treat every comp band, date, and “typical at this level” claim as a hypothesis to verify.
Prompt anatomy
Every JD-fit prompt should carry six elements:
- Role: which persona AI plays — candidate, hiring manager, or recruiter.
- Context: target role, company, level, and your background.
- Goal: one deliverable — a table, a score, a decision.
- Constraints: word count, banned phrases, facts that must appear.
- Tone: confident / curious / measured — pick 2-3 anchors.
- Examples: paste 1-2 of your past bullets so the model matches your record, not a template.
A note on the placeholders below: replace [jd], [me], and [title] with your own pasted text. Keep the bracketed-square style — it signals to the model “fill this from what I give you,” which reduces invented detail.
12 copy-ready prompt templates
1. Real vs wishlist requirements
JD: [jd]. Identify: (a) MUST-haves (the role won't function without them), (b) NICE-to-haves (wishlist), (c) cargo-cult lines copied from another posting. For each line, say how you can tell which bucket it belongs in. Output a 3-column table.Swap: [jd]
2. Where I fit / don’t
JD: [jd]. My background: [me]. Score fit on 7 dimensions: domain, level, skills, scope, leadership, cultural, comp range. Each 1-5 + a one-line rationale. Total >= 25 = strong apply; 18-24 = stretch; under 18 = pass.Swap: [jd], [me]
3. Stretch role analysis
I'm a stretch for this role. Identify: (1) 2 things I genuinely lack, (2) 2 ways I can credibly compensate, (3) 1 area I should not pretend to have. Be blunt — pretending gets exposed by the third interview.4. Hidden signals
Read between the lines of this JD: (1) Why does this role exist now — new team, backfill, or expansion? (2) What scope creep is hinted at? (3) What culture clues do the verbs leak ("thrive in ambiguity", "self-starter", "wear many hats")? Output a short between-the-lines note.5. Title-vs-role mismatch
The JD title says [title]. Read the responsibilities. Does this role actually match the title at typical companies? Flag whether it looks (a) under-leveled, (b) over-leveled, or (c) unusually scoped, and tell me how to adjust my expectations.Swap: [title]
6. Apply / pivot / pass decision
Based on my fit analysis above, decide: (a) APPLY now, (b) PIVOT (same kind of role but at a smaller / larger / different-stage company), or (c) PASS. Give a one-paragraph rationale plus the next 3 concrete actions.7. Skills to learn before applying
For the genuine gaps in my fit, list: (a) skills I could close meaningfully in 4 weeks, (b) a project I could ship as proof, (c) what NOT to invest in because it won't move the decision. Be ROI-aware and rank by impact.8. JD red flags
Audit this JD for red flags: "wears many hats" with no team, "must be passionate" with no comp, "fast-paced" with no leadership context, an unrealistically long requirements list, or a title/responsibility mismatch. Output each flag plus its impact on whether I apply.9. Compensation signal
From this JD infer: company size, location, role level, equity language. Estimate a likely comp band and explain the reasoning. Cross-check against levels.fyi and public salary data, footnote your uncertainty, and do not invent a specific offer number.10. Cover-letter anchor extraction
From this JD, extract the 3 phrases I should echo (not parrot) in a cover letter — the specific responsibility or value they prioritise. For each, write one sentence linking it to my real background: [me].Swap: [me]
11. Recruiter screen prep
Based on my fit analysis, prep recruiter-screen talking points: (a) a 30-second elevator opener, (b) one example matching their top requirement, (c) the one honest gap question they'll likely ask and how I'd answer, (d) 2 sharp questions I should ask them.12. JD diff across 3 companies
Compare 3 similar JDs: [jd1], [jd2], [jd3]. For each: scope, level signal, must-haves, culture clues. Then recommend which one to prioritise for application and why.Swap: [jd1], [jd2], [jd3]
Common mistakes
- No specific context (company / role / level). Generic input gets generic output.
- Asking for honesty without your record. With nothing to anchor to, the model confabulates strengths and gaps.
- Trusting numbers it hands back. AI invents comp bands, dates, and titles. The 9. compensation prompt exists to force a source check, not to replace one.
- One answer reused across companies. Recruiters at the same firm compare notes; identical phrasing reads as a blast.
- No tone anchor. Without 2-3 tone words the output lands flat.
- Treating the first draft as final. Raw model output reads AI-flavoured; human-edited bullets scored about 27% higher on job-match relevance in 2026 testing.
How to push results further
- Paste real bullets so the model matches your voice, not a stock template.
- Run #2 (fit score) and #6 (decision) as a pair, then have the model argue the opposite call to stress-test it.
- For stretch roles, ask AI to play the interviewer first — weak answers surface before a real one does.
- Keep one doc per company: the fit analysis, gaps to address, and questions to ask.
- Re-run a quick scan the morning of the interview for news or launches in the past week.
FAQ
- Can recruiters tell when an application is AI-written? Increasingly, yes — not because of detectors but because low-effort AI output is now generic and everywhere. Specifics from your real record are the antidote.
- Which model should I use for fit analysis? Claude Opus 4.7 or Sonnet 4.6 if you want it to flag missing data instead of guessing; GPT-5.5 if you’re pasting a JD screenshot. Verify numbers either way.
- How much research is enough before applying? For a role you genuinely want, 60-90 minutes including the prompts above. Past that, returns diminish fast.
- When should I start salary research? Before you apply, using the #9 prompt as a starting hypothesis. Validate against 2-3 sources (levels.fyi, Glassdoor, recent postings) before any negotiation.
- Does tailoring actually change my odds? Cold applications convert at roughly 2-3% to interview; tailored, well-fit applications run materially higher. The triage here is what tells you which roles are worth that tailoring time.
- Should I apply to stretch roles anyway? Often yes — but use prompt #3 to name what you lack honestly, so you can answer for it rather than be exposed by it.
Related
- JD matching prompts
- Resume prompts
- Cover letter prompts
- Recruiter reply prompts
- AI Job Description Analysis: Must-Haves, Gaps, Likely Questions
- Career & Interview Prompts hub
Tags: #Prompt #Job search #JD analysis