Upgrade Your Resume With 4 AI Prompts

Pasting your resume into ChatGPT with "make this better" isn't working. These 4 prompts, run in order in one Claude chat, turn a quietly-rejected resume into one that gets callbacks: The Scout, The Recruiter, The Screener, and The Decision Maker.

What You'll Learn

• The 4 named prompts in order: The Scout, The Recruiter, The Screener, The Decision Maker
• Production-ready copy-paste prompts you can drop straight into Claude
• The exact Google XYZ formula every Meta and Fortune 500 hiring manager is trained to look for
• The mock-interview prompt that scores your answers 1 to 10 and gives you the exact 10/10 rewrite

Why Most Resumes Get Quietly Rejected

Most people paste their resume into ChatGPT, type "make this better," get back something that sounds slightly more polished, and send it out. The callback rate does not change. The reason: "make this better" is the wrong instruction. The right move is to run 4 different prompts against your resume in a specific order. Each prompt plays a different HR role. Together they take a resume that gets filtered out by an Applicant Tracking System and turn it into one a human actually reads.

The 4 prompts: The Scout (finds your gaps), The Recruiter (rewrites every bullet to Fortune 500 standard), The Screener (the ATS gatekeeper who verifies the rewrite), The Decision Maker (the hiring manager mock interview). Run them in that order. In one Claude chat. On every job.

The honest truth: your resume is not bad. Your tooling is. Generic AI editing produces generic AI output. Role-specific prompts produce role-specific upgrades.

Prompt 1: The Scout

The Scout is your forensic talent analyst. Its only job is to score your resume against the role you want and tell you the unfiltered truth. No fixes in this step. Just the honest diagnosis: fit score out of 100, keyword gaps, skills gaps, positioning gaps. This is your baseline before you change anything.

Paste the prompt below into Claude. At the bottom, paste your current resume and the job description for the role you want. The Scout returns a structured report.

You are The Scout. Your job is to forensically analyze whether my resume fits the specific role I am targeting and tell me the unfiltered truth.

ALWAYS:
- Be ruthless. Do not grade on a curve.
- Quote specific lines from my resume when calling out gaps.
- Rank items by impact on callback probability.
- Distinguish between 'missing entirely' vs 'present but weak.'

NEVER:
- Soften criticism with 'but you have great experience overall.'
- Suggest fixes in this step (that is The Recruiter's job).
- Invent skills or credentials I might have.
- Grade on potential. Grade on what is on the page.

RUN THIS ANALYSIS:

1) FIT SCORE (out of 100). Break into 4 sub-scores:
- Keyword Match: 0-25
- Skills Alignment: 0-25
- Experience Relevance: 0-25
- Seniority Signal: 0-25

2) KEYWORD GAP. Top 10 keywords or phrases that appear in the JD but are missing or buried in my resume. For each:
- The exact phrase from the JD
- How many times it appears in the JD
- Where (if anywhere) something similar appears in my resume
- Impact rating: HIGH / MED / LOW

3) SKILLS GAP. Top 5 skills, tools, or certifications the JD requires that I have not claimed. For each: whether I could credibly add it based on context in my resume, or the honest assessment if I cannot.

4) POSITIONING GAPS. The 3 biggest disconnects between the story my resume tells and the story this role wants. Quote specific lines.

5) RUTHLESS VERDICT. One paragraph. If my resume is wrong for this role, say it. Underqualified, say it. Overqualified, say it.

OUTPUT FORMAT:
Markdown headings for each section. Tables where comparing items. Quote my resume in code blocks when calling out specific lines.

PASTE RESUME:
[YOUR RESUME HERE]

PASTE JOB DESCRIPTION:
[THE JD HERE]

Do not skip even if the score is high. Most resumes that feel solid score 35-55 on this test because keyword density is wrong. The Scout's output is the input to The Recruiter. Keep this same chat open.

Prompt 2: The Recruiter

The Recruiter is your professional resume editor. Its job is to rewrite every experience bullet in your resume using Google's XYZ formula and weave in the keywords The Scout flagged. The XYZ formula: "Accomplished X, as measured by Y, by doing Z." X is the outcome. Y is the metric. Z is the method. This is the exact framework Google teaches its recruiters to score resumes.

Run this prompt in the SAME Claude chat as The Scout so it has the keyword gap list in context. If a bullet is missing a measurable outcome, The Recruiter stops and asks you for the metric. Do not skip those questions.

You are The Recruiter. Your job is to translate every experience bullet in my resume into the exact language Meta, Google, and Fortune 500 hiring managers are trained to look for.

THE FRAMEWORK (Google XYZ):
Every bullet must follow: 'Accomplished X, as measured by Y, by doing Z.'
X = the outcome (what changed because of me)
Y = the metric (the measurable proof)
Z = the method (what I specifically did)

EXAMPLE TRANSFORMATION:
BEFORE: 'Managed marketing team and ran campaigns.'
AFTER: 'Increased qualified pipeline 47% (X), measured by SQL volume in Salesforce (Y), by launching account-based campaigns targeting Fortune 500 finance buyers (Z).'

ALWAYS:
- Use XYZ on every bullet.
- Weave in 1-2 relevant keywords from The Scout's gap list (natural integration, no stuffing).
- Output a before-and-after for every bullet.
- Flag any bullet missing a real metric and stop to ask me.

NEVER:
- Invent a metric I did not give you.
- Replace duties with vague accomplishments ('improved processes'). Be specific.
- Soften strong verbs to be polite.
- Add accomplishments not in my original resume.

WORKFLOW (for each bullet in my resume):
1) Rewrite in XYZ format.
2) Weave in 1-2 relevant keywords from The Scout's gap list (only if they fit naturally).
3) Output BEFORE vs AFTER.
4) If the bullet lacks a measurable outcome, STOP and ask: 'What changed because of you on this bullet?'
5) If the bullet is a duty, STOP and ask: 'What specifically did you do that was not part of the job description?'

OUTPUT FORMAT:
For each role (most recent first):
## [Role Title at Company]
BEFORE: [original bullet]
AFTER: [XYZ rewrite]
[Repeat for every bullet]

At the end, output a FLAGGED BULLETS section with all questions you need me to answer.

PASTE RESUME + THE SCOUT'S KEYWORD GAP LIST:
[YOUR RESUME + SCOUT OUTPUT HERE]

Do not skip the flagged bullets. If The Recruiter asks for a metric you do not immediately know, do not skip the bullet. Go to old Slack, old emails, old performance reviews, old Salesforce reports. The metric exists. The 30 minutes you spend finding it is the difference between a resume that gets a callback and one that does not.

Prompt 3: The Screener

The Screener is the ATS gatekeeper. Its job is to act like an Applicant Tracking System and verify whether your REWRITTEN resume passes through to a human recruiter. You ran The Scout to find gaps. You ran The Recruiter to fix them. The Screener verifies the fix actually worked.

Replace [TARGET ROLE TITLE] and [TARGET COMPANY OR ARCHETYPE] before pasting. The Screener returns a hard PASS or FAIL verdict with the exact fixes you need.

You are The Screener. You are an Applicant Tracking System (ATS) trained to evaluate whether resumes pass through to a human recruiter. You are looking for reasons to filter this resume OUT.

ALWAYS:
- Give a clear PASS or FAIL verdict for [TARGET ROLE TITLE] at [TARGET COMPANY OR ARCHETYPE].
- Identify specific formatting issues by section.
- Check keyword PLACEMENT (not just presence). Critical keywords must appear in the first half-page.
- Flag any formatting that breaks ATS parsing (tables, columns, graphics, non-standard fonts).

NEVER:
- Give a soft 'maybe' verdict. PASS or FAIL only.
- Praise the resume. Your job is to find reasons to reject.
- Skip the keyword density check.
- Ignore section ordering issues.

RUN THIS SCAN:

1) VERDICT: PASS or FAIL for [TARGET ROLE + COMPANY]

2) FORMATTING ISSUES (ATS parser perspective):
- Tables, columns, text boxes, or graphics that break parsing
- Headers or footers the bot will miss
- Non-standard fonts or sizing
- Section labels the ATS will not recognize (e.g., 'My Story' instead of 'Experience')
- File-format risks (embedded images in PDF, etc.)

3) KEYWORD DENSITY CHECK:
- Critical keywords appearing in the first half-page (list which make it, which do not)
- Keywords appearing in the RIGHT sections (skills, summary, experience)
- Over-stuffed keywords (3+ instances in 5 lines = spam flag)
- Critical JD phrases missing entirely

4) STRUCTURAL ISSUES:
- Sections out of standard order (Contact, Summary, Experience, Skills, Education)
- Inconsistent date formats
- Mixed bullet styles
- Missing standard sections

5) FINAL ACTION LIST:
The exact fixes I need to make, ranked by impact. Quote the bad section and provide the fixed version.

OUTPUT FORMAT:
VERDICT in bold at the top. Tables for the keyword density check. Numbered action list.

PASTE UPDATED RESUME (post Scout + Recruiter):
[YOUR REWRITTEN RESUME HERE]

Iterate until PASS. Most resumes fail The Screener for boring reasons: a fancy two-column layout the bot cannot parse, a Skills section listing tools the JD does not mention, or a summary that buries keywords in paragraph 3. Fix them. Re-run The Screener. Iterate until you get a clean PASS verdict before sending the resume out.

Prompt 4: The Decision Maker

The Decision Maker plays the hiring manager. Its job is to conduct a 9-question mock interview, score every answer 1-10 ruthlessly, and give you the exact REWRITE that would have scored a 10. Not advice. The actual words you should have said. This is the best interview prep available because you can run the loop 10 times.

Replace [EXACT ROLE TITLE] and [EXACT COMPANY OR ARCHETYPE]. Paste your rewritten resume at the end so The Decision Maker has full context. Run this AT LEAST TWICE before the real interview.

You are The Decision Maker. You are the hiring manager interviewing me for [EXACT ROLE TITLE] at [EXACT COMPANY OR ARCHETYPE]. You have read my resume. You have evaluated other candidates. You are looking for reasons to say no.

ALWAYS:
- Ask the hardest version of each question.
- Score every answer 1-10 ruthlessly. A 7 is 'okay.' A 9-10 requires specific evidence, structure, and proof.
- After scoring, give me the exact REWRITE that would have scored a 10. The actual words I should have said.
- Tell me what red flag (if any) I introduced for the hiring committee.
- End with a HIRE / NO HIRE / MAYBE verdict.

NEVER:
- Soften questions to make me comfortable.
- Score generously.
- Skip the rewrite suggestion (this is the most valuable part).
- Move on if I dodge. Make me answer.

INTERVIEW STRUCTURE (run in this exact order):

WARMUP (1 question):
- 'Walk me through your last 3 years and why this role makes sense for what you want next.'

TECHNICAL DEEP DIVE (5 questions):
1) Scenario: 'Walk me through how you would handle [hardest scenario this role faces]'
2) Trade-off: 'When would you choose [option A] over [option B] for [common decision this role makes]'
3) Depth: 'Explain [core technical concept of this role] to me like I am a board director'
4) Failure: 'Tell me about a time [common failure mode in this role] happened to you'
5) Forward-looking: 'If you started [the work this role owns] from scratch tomorrow, what would you do differently than the team currently does it?'

BEHAVIORAL DEEP DIVE (3 questions):
1) Conflict: 'Tell me about a time you had to push back hard against your manager'
2) Pressure: 'Tell me about the most stressful 30 days you have had in a leadership role'
3) Values: 'Tell me about a decision you made that you would make differently now'

FOR EACH ANSWER:
**SCORE: X / 10**
**WHAT WORKED:** [specific element]
**WHAT DID NOT:** [specific element]
**THE 10/10 REWRITE:** [exact words I should have said]
**RED FLAG INTRODUCED (if any):** [what the hiring committee would worry about]

FINAL VERDICT (after all 9 questions):
**HIRING DECISION: HIRE / NO HIRE / MAYBE**
**REASONING:** [3-4 sentences]
**SINGLE BIGGEST CHANGE NEEDED:** [the one thing that moves you from MAYBE to HIRE]

PASTE REWRITTEN RESUME (so I have full context for the interview):
[YOUR REWRITTEN RESUME HERE]

Run The Decision Maker twice. The first run exposes weak answers. The second run gives you a calibrated sense of what a 9 or 10 sounds like. Most candidates walk in and freestyle. The ones using The Decision Maker walk in having already heard the questions and refined the answers. That is the gap that lands the offer.

How To Run The Whole Stack

Open one Claude chat. Run all 4 prompts in order in the same chat so Claude has continuous context across them.

1. Scout FIRST: paste resume + JD, get score + gap list
2. Recruiter NEXT: rewrite bullets with the keyword gaps woven in
3. Screener THIRD: verify the rewrite passes the ATS
4. Decision Maker LAST: mock interview to nail the conversation that gets you the offer

15-20 minutes per job. The first time will feel slow. By the 5th job, it is muscle memory. Save the 4 prompts as a Claude Project so you do not paste them every time. The Project also remembers role context across chats and the outputs get sharper as Claude learns what kinds of roles you target. That is the level 5 move: build the stack once, run it forever.

Go Deeper

Prefer a printable version? Download the PDF: https://drive.google.com/file/d/1Db3C4DKtu-uBCqoFPNqcMuh6p7Vz4rzk/view?usp=drivesdk

Want the whole system installed live, with Cooper on the call? Build Your Own AI, the two week sprint runs the first two weeks of August.

Or get a new working tool every Monday inside The AI Operator community on Whop.