Hiring & People · Updated July 17, 2026
ChatGPT prompt for candidate screening questions
This is the prompt we run once the job post is live, to build the filter before applications arrive. It turns the scorecard into a full screening kit. Its rule: it adds no criterion you didn’t set; everything comes from the scorecard, nothing else.
The prompt to copy
From this scorecard, build a screening kit:
1. Five application questions that test for the must-have outcomes — each answerable in ≤5 minutes, no riddles or trick questions.
2. For each question: what a strong answer and a weak answer look like, grounded only in the scorecard — don't add criteria I didn't set.
3. A 15-minute phone-screen script: 6 questions in order, each with one follow-up probe.
Scorecard: {paste}
Paste the scorecard from H1, then run it: the kit comes out ready to drop into your application form and your call agenda.
When to use it
The job post is live, applications are about to land. This kit gets built before, so you screen everyone through the same filter instead of improvising CV by CV.
What you get back
- Five application questions answerable in five minutes.
- For each, what a strong answer and a weak answer look like.
- A 15-minute phone-screen script: six ordered questions, each with a follow-up probe.
- A filter grounded in the scorecard, not in gut feel.
Real example: from scorecard to filter
Input: Lena, at the Brightside agency, pastes the project-lead scorecard. Output excerpt:
Question 2: Describe a project you delivered late. What did you change after?
Strong answer: a concrete, measurable follow-up action (tied to the
scorecard outcome "deliver on time").
Weak answer: blames the client, no action of their own.
Notice the expected answer maps to a specific scorecard outcome, not to a criterion pulled from nowhere. The AI doesn’t judge “personality” or “fit,” because you didn’t put those in the scorecard. That is the prompt keeping you honest.
The full screening kit, with the five questions and the phone-screen script, ships inside the kit alongside Playbook P7.
Frequently asked questions
- Does this prompt work with Claude or Gemini?
- Yes. It relies on no ChatGPT-specific feature: paste it as-is into Claude or Gemini and the result is the same.
- Why must the questions test only the scorecard?
- Because screening on unwritten criteria means deciding on gut feel and letting bias in. By grounding only in the scorecard, the kit filters on what you actually defined as must-have.
- Can I use the phone-screen script without the application questions?
- Yes. The prompt produces three independent blocks: five application questions, their expected answers, and a 15-minute phone-screen script. Take whichever fits your screening stage.