Guide
ChatGPT prompts for HR, and where to save them
The HR prompts worth keeping are the ones for writing HR repeats: a job post from a role brief, structured interview questions with a scoring guide, a first-90-days onboarding plan, and answers to the questions employees ask about a policy. All four work without anyone's personal data, which is the first rule for HR and AI. Copy them below, add your templates, and save them where your AI reads them on its own: your style in custom instructions, each open role in a project, each repeated task as a skill.
Short answer
- Four prompts: job post, interview questions, onboarding plan, policy FAQ.
- No personal data in any of them.
- Save them: style in custom instructions, roles in projects, tasks as skills.
- People decide: AI drafts, it doesn't rank candidates.
The four prompts
1. Job post
"Write a job post for [role] from this brief: [paste]. An honest title, what the person will do in the first six months, at most five must-haves, the salary range [range], and how to apply. Plain language, no 'rockstar', no 'fast-paced family'."
2. Structured interview questions
"For [role], write six interview questions tied to these must-haves: [list]. For each, what a strong, an acceptable and a weak answer sound like. Same questions for every candidate."
3. Onboarding plan
"Draft a first-90-days plan for a new [role]: what must be ready before day one, the first week by day, people to meet and why, a first piece of work, and goals for each 30 days. Our tools: [list]."
4. Policy FAQ
"From this policy, write the ten questions employees are most likely to ask, each answered in two plain sentences with the section it comes from. Policy: [paste]."
This one predicts the questions before a policy goes out. Once real questions come in, round them up instead: ChatGPT for HR professionals starts there, with which HR jobs to give ChatGPT first.
Where to save each
| What | Where | Why |
|---|---|---|
| Company style, values, rules | Custom instructions | Applied in every chat |
| One open role's brief and scorecard | A project per role | Instructions and files apply inside the project |
| A task you repeat for every role | A skill | Loaded when you ask for the task |
From OpenAI's and Anthropic's help pages, read 5 November 2026.
A worked example
Illustration: a made-up business. Hana is the only HR person at a 70-person logistics company with eight open roles. She'd been pasting the job-post prompt and a page of style rules for each role.
- Custom instructions: "HR at a logistics company. Plain English. Salary ranges always shown. Never ask for or include candidates' personal details."
- A project per open role with the hiring manager's brief and the scorecard.
- Job post and interview questions as skills, with her template and two past posts that attracted good applicants.
For the warehouse supervisor role she types "job post for this role" inside its project and gets her format with the manager's must-haves. She still reads every line: the first draft listed a forklift licence the manager had marked "nice to have".
Save your first prompt
- Pick the prompt you paste most, usually the job post.
- Add your template and two good examples.
- Move company style and rules to custom instructions.
- Save the task as a skill and test it on two real roles.
- Delete the old copy from your notes.
Next step: HR's rules for everyone
The onboarding plan prompt is the start of the next step: using AI for onboarding new employees covers the rest of the preparation and what stays with the manager. Once hiring managers use AI too, your rules need to reach them. The HR guide to an employee AI policy has sample wording, and teaching a new hire your AI workflow covers day one.
Skills for this
- AI practiceTurn a prompt into a skillTurns a prompt that works into a skill the whole team can reuse, with a valid name, a when-to-use description, steps, an output template, an example, and checks.
- HiringJob post from a role intakeWrites a job post with an honest title, the first 6 months of work, 5 or fewer must-haves, a pay-range line checked for local rules, and an inclusive-language check.
- HiringStructured interview questionsWrites behavioral and situational interview questions mapped to competencies, with prepared probes, strong and weak answers, and a never-ask list.
- Operations and adminEmployee onboarding plan (30/60/90)Drafts a 30/60/90-day plan for a new hire: accounts and tools ready by day 1, week 1, people to meet, a buddy, a first deliverable, goals, and check-ins.
Questions
What are good ChatGPT prompts for HR?
Prompts for the writing HR repeats: a job post from a role brief, structured interview questions with a scoring guide, a first-90-days onboarding plan, and answers to the questions employees ask about a policy. Each works without any personal data.
Can HR paste CVs into ChatGPT?
Keep candidate and employee personal data out of personal AI accounts. On a company plan, follow your rules, and don't let AI rank candidates: some places, such as New York City, regulate automated tools in hiring.
Where should HR save the prompts it reuses?
Your company's style and rules in custom instructions, one role's brief and scorecard in a project, and each repeated task, such as job posts, as a skill.
AI skill
Have your AI turn a prompt that works into a skill your team can reuse
Put this skill in AI Skill Pod and everyone on your team gets it in their Claude, Cursor or ChatGPT (Pro, Business, Enterprise): one invite link, free to start. When one of these HR prompts works, Turn a prompt into a skill has your AI turn it into a skill with your templates and rules, such as no candidate data, so everyone in HR uses your latest version.
14 days free · No card needed
Sources
- OpenAI Help: ChatGPT custom instructions — applied across chats; read through search-result text (read 2026-11-05)
- OpenAI Help: Projects in ChatGPT — read through search-result text (read 2026-11-05)
- Claude Help Center: Using skills in Claude — every plan with code execution on (read 2026-11-05)
- NYC DCWP: Automated Employment Decision Tools (Local Law 144) — bias audit and notice before using such a tool (read 2026-11-05)
- OpenAI Help: Data controls — personal plans train on chats by default unless turned off; read through search-result text (read 2026-11-05)