A ChatGPT resume writer prompt should do three things: use the actual job description, preserve only verified experience and show what information is missing. It should not invent employers, dates, skills or performance numbers.
OpenAI’s prompting guidance recommends clear, specific instructions, enough context and iterative refinement. The template below turns that advice into a resume workflow.
On this page
Prepare the inputs
Gather:
- the full job description;
- your existing resume;
- a list of projects and responsibilities;
- verified outcomes or numbers;
- location and work-authorisation facts you are comfortable including;
- the target length and format.
Remove addresses, phone numbers and other sensitive identifiers from the working copy if they are not needed for drafting.
The prompt
Act as a resume editor. Tailor my existing resume to the job description below.
Rules:
- Do not invent an employer, title, date, degree, tool, certification, responsibility or result.
- Use only facts in “My experience.”
- If the job asks for evidence I have not supplied, list it under “Missing evidence” instead of adding it.
- Keep wording plain and specific. Avoid “results-driven,” “dynamic,” “passionate” and unsupported superlatives.
- Write a two-line summary and up to five bullets per relevant role.
- Start each bullet with the work performed, then the outcome. Use a number only when my source includes it.
- Preserve truthful seniority and scope.
Output:
- Match map: job requirement → evidence from my experience
- Resume draft
- Missing evidence
- Claims I should verify before sending
Job description: “”" [paste job description] “”"
My experience: “”" [paste existing resume and verified notes] “”"
The separators make the instruction and source material easy to distinguish.
Ask for a match map first
The match map shows why each change is being made. It also exposes gaps before they become invented bullets.
If a requirement has no supporting evidence, leave it out of the resume or add a truthful adjacent skill. Do not convert “worked with the sales team” into “led enterprise sales strategy.”
Improve weak bullets
A weak bullet names a duty:
Responsible for monthly reporting.
A stronger factual bullet explains the work and a verified outcome:
Built the monthly revenue report in Excel and reduced the closing review from three days to two.
If you cannot support the time reduction, remove it:
Built and maintained the monthly revenue report in Excel for finance review.
Specific work is better than a fabricated metric.
Tailor keywords without copying
Use the employer’s terminology when it accurately matches your experience. Do not copy whole sentences from the job description or add tools you have never used.
Ask ChatGPT to highlight every added keyword and point to the source line in your experience. Then review each match yourself.
A useful second prompt
After the first draft:
Audit this resume against my source notes. Create a table with every factual claim, the source passage that supports it and a risk label: supported, ambiguous or unsupported. Do not rewrite yet.
This separates verification from polishing.
What ChatGPT cannot verify for you
It cannot know whether a job title, date, metric or responsibility is true unless you provide reliable records. It also cannot predict an applicant tracking system’s decision.
Treat the draft as editing assistance. Check dates against contracts or profiles, metrics against reports and skill claims against work you can discuss in an interview.
