What is in this pack

This page covers chatgpt prompts for human resources teams: ready-to-use prompts for recruiting, onboarding, performance management, policy writing, and employee communications. It is designed for HR managers and team leads who want consistent, professional output without having to write prompts from scratch every time.

The pack contains 20+ prompts organized into five categories. Each one includes a short note on when to use it and which variable to change first so you can adapt it to your context quickly.

One honest caveat before you dig in: AI is genuinely useful for drafting and structuring HR content, but it should not be the final word on anything that touches employment law, disciplinary action, or compensation. Run those outputs past your legal or HR leadership team before they go anywhere.

For prompts built around other roles, see the full prompts by role library or compare with AI prompts for business teams.

The prompts, by category

These prompts are grouped by the HR task they serve. Each one works in ChatGPT, Claude, Copilot, or Gemini without modification. Variables in [BRACKETS] are the only parts you need to change before running.

Recruiting and job descriptions

Write a job description from a hiring manager's rough notes

When to use: You have a Slack message or a bullet list from the hiring manager and need a complete, postable JD. What to change first: Replace [ROLE] and [TEAM] with specifics, then adjust the seniority signals in the responsibilities section.

Role: [ROLE]
Team: [TEAM]
Seniority: [LEVEL, e.g. mid-level / senior]
Key responsibilities (rough notes): [PASTE NOTES]
Must-have skills: [LIST]
Nice-to-have skills: [LIST]
Company tone (e.g. formal / conversational): [TONE]

Write a complete job description for this role. Include: a two-sentence company intro, a responsibilities section (six to eight bullets), a requirements section split into must-have and nice-to-have, and a closing paragraph on what makes this team worth joining. Keep sentences short. Avoid jargon.

Rewrite a job description to reduce gendered language

When to use: Before posting any existing JD, especially for technical or leadership roles where gender-coded language is common. What to change first: Paste the full existing JD into [EXISTING JD].

Existing job description:
[EXISTING JD]

Rewrite this job description to remove gendered language and unnecessarily exclusionary phrasing. Use research on inclusive job postings as a guide (for example, replacing "competitive" with "driven", "ninja" with "specialist"). Do not change the core requirements or responsibilities. Flag any phrases you changed and explain why in a brief note at the end.

[NEEDS REAL OUTPUT]

Write five Boolean search strings for a hard-to-fill role

When to use: When sourcing on LinkedIn Recruiter or a CV database and generic searches return noise. What to change first: Fill in [ROLE], [MUST HAVE SKILLS], and [EXCLUDE] before running.

Role: [ROLE]
Must-have skills or titles: [MUST HAVE SKILLS]
Exclude: [EXCLUDE, e.g. agencies / staffing firms]
Location: [LOCATION]

Write five distinct Boolean search strings I can use on LinkedIn Recruiter to find candidates for this role. Vary the approach across the five: title-focused, skills-focused, company-type-focused, and so on. Format each string as a copyable line.

Draft a candidate outreach message for a passive candidate

When to use: Cold outreach via LinkedIn InMail or email, where the candidate has not applied. What to change first: Add something specific about the candidate in [CANDIDATE DETAIL] so the message does not read as a mass send.

Role I am hiring for: [ROLE]
Candidate detail (something specific to them): [CANDIDATE DETAIL]
What makes this role or company worth considering: [VALUE PROPOSITION]
Desired tone: [TONE, e.g. warm and direct / professional]
Word limit: 100 words or fewer

Write a personalized outreach message to a passive candidate. Lead with the specific detail about them. Do not open with "I came across your profile."

Generate structured interview questions for a role

When to use: Building an interview guide before a hiring loop kicks off. What to change first: Specify [COMPETENCIES] based on what you actually need to test, not a generic list.

Role: [ROLE]
Competencies to assess: [COMPETENCIES, e.g. stakeholder management, prioritization under pressure]
Interview format: [FORMAT, e.g. 45-minute panel / two-stage behavioral]

Write a structured interview guide with four to six questions per competency. For each question, include: the question itself, what a strong answer looks like, and one follow-up probe. Use the STAR format as the basis but do not instruct the interviewer to explain STAR to the candidate.

[NEEDS REAL OUTPUT]

Onboarding

Write a 30-60-90 day plan for a new hire

When to use: Before a new hire's first week, in partnership with their manager. What to change first: Replace [ROLE] and [TEAM CONTEXT] with real specifics. Vague inputs produce vague plans.

New hire role: [ROLE]
Team context: [TEAM CONTEXT]
Key priorities in the first quarter: [PRIORITIES]
Tools and systems they will use: [TOOLS]

Write a 30-60-90 day onboarding plan. Structure it in three phases. Each phase should include: learning objectives, key relationships to build, milestones to hit, and one definition of success by the end of the phase. Keep it practical. Avoid corporate filler.

Draft a welcome email from the hiring manager

When to use: The day before or the morning of a new hire's start date. What to change first: Add the manager's actual voice notes in [MANAGER NOTES] so it does not read as a template.

New hire name: [NAME]
Role: [ROLE]
Start date: [DATE]
Manager name and role: [MANAGER NAME AND ROLE]
Manager's personal notes or things they want to say: [MANAGER NOTES]
First day logistics (where to go, who to ask for, etc.): [LOGISTICS]

Write a warm, practical welcome email from the hiring manager to the new hire. Open with something personal. Include the first-day logistics clearly in the second half. Keep it under 200 words.

Create an onboarding checklist for a remote employee

When to use: When your standard checklist was built for in-office hires and a remote joiner has different needs. What to change first: Adjust [TOOLS AND SYSTEMS] to match your actual tech stack.

Role: [ROLE]
Location: [CITY OR COUNTRY]
Tools and systems: [TOOLS AND SYSTEMS]
Key contacts: [KEY CONTACTS]

Create a first-week onboarding checklist for a remote employee. Separate tasks by owner: HR, IT, manager, and new hire. Include async-friendly options for any activity that would typically be in-person. Flag any tasks that need to be completed before day one.

Performance and feedback

Turn bullet-point feedback notes into a written review

When to use: Managers often have good instincts but write reviews that are vague or legally risky. This prompt bridges the gap. What to change first: Paste real notes into [MANAGER NOTES]. Vague input produces vague output.

Employee role and level: [ROLE AND LEVEL]
Review period: [PERIOD]
Manager's rough notes: [MANAGER NOTES]
Rating or overall assessment: [RATING, if applicable]
Tone guidance: [TONE, e.g. direct / developmental / formal]

Rewrite these notes as a structured performance review. Organize into: strengths (two to three), areas for development (one to two), and a summary paragraph. Use specific language. Avoid phrases like "good team player" or "hard worker" without concrete examples. Do not add examples that are not in the notes.

[NEEDS REAL OUTPUT]

Write a performance improvement plan (PIP) outline

When to use: When a formal PIP is needed and you want a consistent, documented structure. What to change first: Fill in [PERFORMANCE GAPS] carefully. This document may be used in an employment proceeding.

Employee role: [ROLE]
Performance gaps: [PERFORMANCE GAPS]
Expectations not being met: [EXPECTATIONS]
Support being offered: [SUPPORT]
Timeline: [TIMELINE, e.g. 60 days]
Review checkpoints: [CHECKPOINTS]

Draft a Performance Improvement Plan outline. Include: a statement of the issue, the specific standards expected, the support the company will provide, the timeline and review dates, and the consequences if performance does not improve. Write in plain, professional language. Do not editorialize or assign intent to the employee.

Generate upward feedback questions for a 360 review cycle

When to use: When asking employees to give feedback on their managers and you need questions that surface real information, not just satisfaction scores. What to change first: Adjust [FOCUS AREAS] if the review cycle has a particular theme (e.g. post-restructure, new manager).

Manager role: [MANAGER ROLE]
Focus areas for this cycle: [FOCUS AREAS, e.g. communication, decision-making, team development]
Number of questions: [NUMBER, e.g. 8]
Response format: [FORMAT, e.g. open text / 1-5 scale with comment]

Write upward feedback questions for a 360 review. Make questions specific enough to produce useful answers. Avoid leading questions. Include at least two questions that give employees space to name something the manager does well and at least two that invite constructive input.

Draft a calibration talking-point sheet for a manager

When to use: Before a calibration session, to help managers prepare consistent, evidence-based positions on their team members. What to change first: Replace [EMPLOYEE SUMMARIES] with brief notes on each employee rather than full reviews.

Team members being calibrated: [NAMES AND ROLES]
Employee summaries (one paragraph each): [EMPLOYEE SUMMARIES]
Rating scale used: [RATING SCALE]
Known tensions or edge cases: [TENSIONS, if any]

Write a calibration talking-point sheet for the manager. For each employee, include: a one-sentence position statement, two to three pieces of supporting evidence, and a likely challenge from other reviewers with a suggested response. Keep each section brief enough to reference quickly in a live meeting.

[NEEDS REAL OUTPUT]

Policy and documentation

Summarize a policy document in plain language

When to use: When employees are ignoring long policy documents and you need a readable summary to post on an intranet or send by email. What to change first: Paste the full policy into [POLICY TEXT]. Do not summarize a summary.

Policy document: [POLICY TEXT]
Audience: [AUDIENCE, e.g. all staff / managers only]
Format needed: [FORMAT, e.g. one-page summary / FAQ / bullet list]

Summarize this policy in plain language for [AUDIENCE]. Use short sentences. Highlight the three to five things employees most need to know. Flag any section where the original language is ambiguous or where employees are most likely to make mistakes.

Draft a policy update announcement

When to use: When a policy has changed and you need a company-wide communication that is clear but does not cause alarm. What to change first: State [WHAT CHANGED] precisely. Vague announcements generate a flood of follow-up questions.

Policy name: [POLICY NAME]
What changed: [WHAT CHANGED]
Why it changed: [REASON]
Effective date: [DATE]
Where employees can find the full policy: [LINK OR LOCATION]
Tone: [TONE, e.g. matter-of-fact / warm]

Write a policy update announcement for all staff. Open with what changed and when. Explain why briefly. Include a clear call to action (what employees need to do, if anything). Close with a contact for questions. Keep it under 150 words.

Create an FAQ document from a new policy

When to use: For any policy that involves a behavior change, especially around leave, pay, or flexible working. What to change first: Add the full policy text to [POLICY TEXT] and seed [LIKELY QUESTIONS] from actual employee queries if you have them.

Policy: [POLICY TEXT]
Likely questions employees will ask: [LIKELY QUESTIONS, if known]
Tone: [TONE]

Read this policy and write a FAQ document that anticipates the questions employees are most likely to have. Write questions in plain language, as an employee would ask them. Answer each question in two to four sentences. Flag any question where the policy does not give a clear answer.

Employee communications

Write an all-hands update from a senior leader

When to use: Quarterly or ad hoc all-hands where the leader has given you talking points or notes but needs a polished script. What to change first: Add the leader's actual voice in [TALKING POINTS]. A flat briefing document produces a flat script.

Leader name and role: [LEADER NAME AND ROLE]
Key messages: [TALKING POINTS]
Context (e.g. post-restructure, end of year): [CONTEXT]
Tone: [TONE, e.g. direct and human / formal]
Length: [LENGTH, e.g. five minutes spoken / 700 words]

Write an all-hands address from this leader. Open with something that orients the audience quickly. Develop the key messages in a logical order. Close with something specific, not a generic motivational statement. Write in the leader's voice, not corporate boilerplate.

Draft a sensitive communication about a restructure or layoff

When to use: When you need to inform employees about a reduction in force and the message needs to be honest, clear, and legally reviewed before sending. What to change first: Review the output carefully with your legal counsel before sending. This prompt produces a starting draft, not a final document.

What is happening: [DESCRIPTION OF RESTRUCTURE OR LAYOFF]
Who is affected: [AFFECTED GROUPS]
Timeline: [TIMELINE]
Support being offered: [SUPPORT, e.g. severance, outplacement]
What employees should do next: [NEXT STEPS]
Tone: Direct and humane. Do not soften the facts.

Draft an employee communication about this restructure. State what is happening clearly in the first paragraph. Do not bury the news. Explain the reason briefly and honestly. Describe the support available. State next steps. Do not use phrases like "exciting new chapter" or "difficult but necessary". Flag any sentence where the language is evasive.

Write an employee engagement survey with open-text questions

When to use: When your annual survey is too long and too closed-ended, and you want richer qualitative data. What to change first: Limit [THEMES] to three or four real priorities or the survey becomes unfocused.

Company context: [CONTEXT]
Themes to explore: [THEMES, e.g. manager effectiveness, psychological safety, career development]
Number of questions: [NUMBER, e.g. 10]
Format: [FORMAT, e.g. mix of scale and open text]

Write an employee engagement survey. Include a mix of rating-scale questions and open-text questions. For open-text questions, write prompts that invite specific answers rather than general sentiment. Avoid double-barreled questions. Keep the total reading time under five minutes.

[NEEDS REAL OUTPUT]

Learning and development

Build a learning path for a new people manager

When to use: When someone is being promoted into their first management role and needs a structured development plan, not just a reading list. What to change first: Adjust [TIMEFRAME] to match your typical promotion-to-onboarding gap and [COMPANY CONTEXT] to reflect your management culture.

New manager name (optional): [NAME]
Role: [ROLE]
Team size: [TEAM SIZE]
Timeframe for the learning path: [TIMEFRAME, e.g. first 90 days]
Company context: [COMPANY CONTEXT, e.g. fast-growth startup / established enterprise]
Resources available: [RESOURCES, e.g. internal L&D, LinkedIn Learning, budget for coaching]

Build a structured learning path for a first-time people manager. Include: a phased structure, the core skills to develop in each phase, suggested resources or activities for each skill, and one practical exercise per phase that the manager can do with their team. Do not pad with generic management theory. Keep it grounded in day-to-day situations.

Write a skills gap analysis framework for a team

When to use: Before a mid-year or annual planning cycle, when you need to identify development priorities across a function. What to change first: Replace [TEAM FUNCTION] and [STRATEGIC PRIORITIES] with what is actually true for this team. Generic inputs produce generic outputs.

Team function: [TEAM FUNCTION]
Current team size: [SIZE]
Strategic priorities for the next 12 months: [STRATEGIC PRIORITIES]
Known skill gaps (if any): [KNOWN

## Real example outputs

Five prompts in this pack have been run against real models. The outputs below are placeholders until those runs are captured and reviewed.

### Job description, before and after

[NEEDS REAL OUTPUT]

**What to notice:** Whether the model removed vague requirements (like "fast-paced environment") and replaced them with specific responsibilities. Good output reads like a document a candidate would actually use to decide whether to apply.

### Offer letter, tone adjusted

[NEEDS REAL OUTPUT]

**What to notice:** The shift in warmth between a formal and a conversational version. The structure should stay consistent; only the register should change.

### Interview questions, role-specific

[NEEDS REAL OUTPUT]

**What to notice:** Whether the questions probe for actual evidence rather than hypothetical answers. "Tell me about a time when..." beats "How would you handle..."

### Performance review, first draft

[NEEDS REAL OUTPUT]

**What to notice:** How the model handles thin input. If you give it one vague sentence, the output will reflect that. This is a feature, not a flaw.

### Policy FAQ, plain-language version

[NEEDS REAL OUTPUT]

**What to notice:** Whether legal language has been genuinely simplified or just paraphrased. Run the output past whoever owns compliance before publishing it.

## How to adapt these for your team

The prompts above are starting points. Most will need two or three small changes before they fit your organization.

The most common adjustment is terminology. If your company calls them "team members" rather than "employees," or uses a specific grading system like "Band 3" instead of "mid-level," swap those terms in before the first run. The model will mirror whatever language you give it.

The second adjustment is tone. A startup hiring its first ops manager needs a different register than a regional hospital filling a compliance role. Add a sentence like "Our company voice is [direct and informal / formal and precise]" near the top of any prompt that produces external-facing copy.

For recurring tasks, such as weekly pulse summaries or monthly interview debrief templates, save a finalized version of the prompt in a [shared team folder](/product/shared-folders) so everyone runs the same baseline. That eliminates the drift that happens when five people each make their own tweaks.

:::keypoint heading="Worth pausing on"
A prompt that works well for one role often breaks for another. Test on a low-stakes task before rolling it out to a full hiring cycle.
:::

One thing this approach does not handle well: anything that requires live data. Compensation benchmarking, headcount forecasting, and compliance checks all need verified numbers from your HRIS or a current salary survey before a model output means anything.

## Save this folder to your workspace

If your team uses these prompts regularly, the most common failure mode is version drift: one person improves a prompt, nobody else finds out, and within a month you have four slightly different job-posting templates scattered across email threads and personal notes.

Convergence's [shared folders for teams](/product/shared-folders) solve that directly. Save this whole pack as a folder, assign it to your HR workspace, and every team member gets the same prompt set across ChatGPT, Claude, Copilot, and Gemini without copy-pasting anything.

:::cta heading="Want your whole HR team running from the same prompts?"
Save this pack to a shared folder in Convergence and every team member gets instant access across whichever AI tool they already use.
:::

A few practical tips before you do:

- Rename the prompts to match your internal terminology. "Hiring manager" means different things in different organizations.
- Lock the prompts that have gone through legal or compliance review, so edited versions don't quietly replace approved ones.
- Add a "last reviewed" note inside any prompt that references company policy, because policies change.

## Frequently asked questions

### Are these chatgpt prompts for human resources compatible with other AI tools?

Yes. Every prompt on this page works in ChatGPT, Claude, Copilot, and Gemini without modification. The variables in brackets are the only parts that need changing regardless of which tool you use.

### Can AI write a full job description without any human review?

No, and you should not let it. AI drafts miss internal salary bands, legal language specific to your jurisdiction, and the tone your employer brand has built over time. Treat every output as a first draft that a human signs off on.

### How do I stop AI outputs from sounding generic?

Add specifics to the prompt. Include your company name, team size, industry, and one concrete detail about the role or situation. Generic inputs produce generic outputs.

### Is this suitable for a solo HR professional or only for teams?

Both. A solo HR generalist covering the full employment lifecycle gets as much value here as a team of ten. The prompts do not assume any particular team structure.

### How do I share these prompts with my whole HR team?

[Convergence shared folders](/product/shared-folders) let you store, organize, and share prompts across a team so everyone pulls from the same tested versions.