Why this matters
Most job-search AI output reads like bad catalog copy: inflated verbs, invented metrics, empty adjectives. Claude, out of the box, resists that a bit more than most models. It pushes back when a prompt asks it to lie, it holds voice better on longer drafts, and it is willing to ask clarifying questions. That makes it a good tool for the parts of job search where nuance matters most: summaries, cover letters, mock interviews, and salary conversations.
This guide is a collection of 30 prompts you can paste into Claude and adapt in under a minute. They are grouped into 6 categories: resume, cover letter, interview prep, negotiation, LinkedIn, and outreach. Each prompt is structured so Claude has a clear role, clear inputs, and a clear output format. That structure is what separates a useful reply from generic slop.
You do not need a paid account to get value from most of these. The free tier of Claude.ai handles all 30 prompts in this guide. If you are using Claude inside a builder (Cursor, Zed, Claude Code), the same prompts work, though you may need to paste your resume or JD once at the top of the conversation.
How to pass context to Claude
The single biggest mistake people make with AI for job search is dumping a resume and a JD and saying "help me." The output reflects the input. Claude does better with structure. A few rules of thumb:
- Paste the full JD, not a summary. A summary loses the exact phrasing recruiters and ATS care about.
- Paste your resume as plain text, not a PDF link. Claude cannot browse your Dropbox; paste the content.
- State the role and seniority up front. "I am applying for a Senior Backend Engineer role at a Series B fintech" sharpens every reply.
- Name the output format. "Return a 3-column table" or "Return 180 words only" constrains Claude into something you can actually use.
- Ask it to flag gaps. Say "Do not invent metrics; if a number is missing, ask me before filling in." This keeps output honest.
A sensible opener for any Claude job-search conversation: "I am [role, YOE]. Target role: [paste JD]. My current resume: [paste]. I will ask specific questions. Do not add fabricated numbers; if a claim needs a number I have not given, ask me."
Resume prompts (1 to 5)
Use these after you have a draft resume. Claude is better at refining than at writing from zero.
1. Rewrite a weak bullet
Here is a bullet from my resume: "Responsible for handling customer queries." Rewrite it using the Action + Context + Result pattern. Keep it under 22 words. If a number is missing, ask me one targeted question before rewriting; do not invent metrics.
2. Polish a summary
Below is my 3-line resume summary. Keep the facts exactly as stated, but tighten the prose, remove filler, and ensure each sentence adds a new claim (role, proof, direction). Output the final version only, then a one-line list of what you changed.
3. Tailor to a JD
I will paste a job description and then my current resume. Return a table with 3 columns: JD requirement, closest match on my resume, suggested edit. Do not fabricate experience; if a requirement has no match, say "gap" in column 2.
4. Skill coverage audit
Read my resume (pasted below). List skills the resume claims explicitly, skills it implies but does not state, and skills absent. Flag any skill that appears in the summary but is not evidenced in any bullet.
5. ATS keyword gap
Compare the JD and my resume. Return 3 lists: (a) exact keywords in both, (b) JD keywords missing from my resume that I could honestly add, (c) JD keywords I should not add because they are outside my real experience.
Cover letter prompts (6 to 10)
This is where Claude earns its keep. Voice and restraint matter more than keywords, and Claude tends to produce less generic prose than competing models when prompted well.
6. Cover letter draft with voice
Write a 250-word cover letter for the role pasted below. Voice: warm, direct, specific. No openings like "I am writing to." Open with a sentence that shows I understand a specific problem the team has. Include one concrete proof from my resume. Close with a 1-sentence ask.
7. Cover letter from resume
Here is my resume and the JD. Draft a cover letter that does not repeat the resume. Pick the single strongest proof that maps to the top JD requirement and expand that story. Everything else is context.
8. Shorten without losing signal
Cut my cover letter to 180 words. Preserve the opening hook, the one proof, and the ask. Remove every hedging phrase ("I believe," "I think," "I feel"). Show before and after word count.
9. Tone match to company
The company culture page reads [paste]. The JD reads [paste]. Rewrite the tone of my draft cover letter to match that voice without changing the facts. Output the rewrite only.
10. Why this company, honestly
I am interested in this company because [reason]. Rewrite that into a 2-sentence paragraph for my cover letter that sounds genuine, not rehearsed. Avoid the words "passionate," "excited," and "thrilled."
Interview prep prompts (11 to 15)
Claude will hold a back-and-forth mock interview without breaking character. Use voice dictation on your phone if you want to rehearse speaking the answers.
11. Mock interview (behavioural)
Act as a senior engineering manager interviewing me for [role]. Ask one behavioural question at a time, wait for my answer, then grade it using STAR on a 1 to 5 scale with a single reason. Start with "Tell me about a time you disagreed with a teammate."
12. STAR story drafting
I have a story: [paste]. Structure it in STAR (Situation, Task, Action, Result). Keep each section to 2 sentences maximum. Flag if the Result has no number, and ask me for one.
13. Pressure-test my answer
Here is my answer to "Tell me about yourself": [paste]. Attack it. What would a skeptical interviewer probe? List the 5 most likely follow-up questions and the weakest sentence in my answer.
14. Technical concept check
I have an interview for [role]. Quiz me on [topic: eg, database indexing]. Ask me 5 questions, one at a time, increasing in difficulty. Grade each answer. After all 5, rank my weakest area and suggest one resource.
15. Reverse interview questions
I am interviewing at [company] for [role]. Draft 6 questions I should ask the panel. 2 for the hiring manager, 2 for a peer engineer, 2 for the skip-level. Avoid generic questions; each should signal I have done my research.
Negotiation prompts (16 to 20)
Salary conversations are where people freeze. Draft the email in Claude first; you will send something 2 notches more confident than what you would have typed cold.
16. Counter-offer script
I received [offer: 18 LPA base]. My target is 22 LPA. Draft an email counter that anchors on market data, names one comparable offer if I have one, and leaves space for them to respond. Tone: warm, confident, unapologetic.
17. Handle "What are your expectations?"
Recruiter asked for my expected salary. My research says the band is 20 to 28 LPA. Draft 3 response options: (a) deflect to band, (b) name my number with justification, (c) ask for their budget. Explain when to use each.
18. Non-cash asks
The comp is capped but they want me. List 10 non-cash items I could negotiate (eg, sign-on bonus, stock refresh, remote days, title). Rank by typical flexibility. Draft how to ask for the top 3.
19. Multi-offer leverage
I have 2 offers: [A: 20 LPA, dream role] and [B: 26 LPA, decent role]. Help me draft an email to A that uses B as honest leverage without sounding transactional. The goal is to close the gap, not to play them.
20. Decline without burning bridges
I am declining this offer because I took another role. Draft a short, warm decline email that keeps the door open for future roles. No false reasons; be direct but gracious. Under 100 words.
LinkedIn prompts (21 to 25)
21. Rewrite my headline
My current LinkedIn headline is [paste]. My target roles are [list]. Rewrite it in 3 variants: (a) keyword-heavy, (b) value-proposition, (c) personality-led. Explain which best fits recruiter search vs founder DMs.
22. About section from resume
Using only the facts in the resume I just pasted, write a LinkedIn About section in first person. 3 paragraphs: what I do, the shape of my experience, what I am looking for. Use "I" without ever starting a sentence with it more than twice.
23. Experience bullet refresh
Here is my current LinkedIn Experience entry for [role]. Rewrite each bullet to be slightly more narrative than a resume bullet. Keep the numbers, but one or two bullets can read as short stories. Under 400 characters each.
24. Post to announce a job search
I am opening up to new roles. Draft a LinkedIn post: 1 hook line, a short context paragraph, 3 bullets of what I am looking for, and a call to action. No hashtags spam; max 3. Tone: direct, not desperate.
25. Engagement without cringe
I want to comment thoughtfully on posts from recruiters in [industry]. Give me 5 comment templates that add a real perspective rather than "Great post!" Each should be 2 to 3 sentences and invite reply.
Outreach prompts (26 to 30)
Cold email is a numbers game, but the opening line decides whether you are in the game at all. Keep every message specific.
26. Cold email to hiring manager
I found a role at [company]. Draft a 120-word cold email to the hiring manager. Subject line that is specific, not clickbait. Open with a line that shows I researched their work. One proof from my resume. A low-friction ask.
27. Referral ask, warm
I know [person] loosely from [context]. I want to ask for a referral to [role at company]. Draft a short message (under 90 words) that respects their time, makes the ask clear, and gives them an easy out.
28. Follow-up after no reply
I emailed [contact] 7 days ago about [role]. No reply. Draft a single 60-word follow-up that adds one new piece of value (a relevant project I shipped, a question, or a small piece of research). No guilt trips.
29. Recruiter InMail reply
A recruiter sent me this InMail: [paste]. I am open but want to qualify before investing time. Draft a 5-line reply that thanks them, asks 3 clarifying questions (comp band, location, stage), and signals real interest without over-committing.
30. Networking coffee request
I want a 20-minute coffee with [person] at [company]. I am not asking for a job. Draft a message that makes clear I want to learn about their path. Offer 3 time slots. Under 100 words.
Claude vs ChatGPT for job hunting
Both work. The differences are at the margins, and those margins matter for nuanced writing.
- Voice consistency. Claude tends to hold a tone (warm, direct, formal) across a long draft. ChatGPT drifts toward its default cheerful register unless you re-anchor.
- Honest gaps. Claude is more willing to say "I do not have enough information; what is the number?" rather than invent one. You still have to prompt for it.
- Pushback. When you ask Claude to review a flawed plan, it will disagree more readily. Useful for negotiation strategy, not useful if you want cheerleading.
- Speed and tools. ChatGPT wins on integrations (images, code execution) out of the box. Claude catches up via projects and artifacts.
For resume and cover letter work specifically, most people who try both prefer Claude output. For research-heavy tasks (company teardowns, market comps), ChatGPT with web browsing is still the faster default.
External references
Frequently asked questions
Do I need a paid Claude account for these prompts?+
Can Claude see my PDF resume?+
Is it safe to share my resume with Claude?+
How do I stop Claude from inventing metrics?+
Does Claude write in first or third person for resumes?+
How do I make Claude sound less like AI?+
Can I use Claude for technical interviews?+
Skip the prompt engineering
ResumeBuildz applies the same quality bar these prompts aim for: no invented numbers, no fluff, clean ATS-ready output. Built-in.