AI prompts for recruiters is a ready-to-use prompt library covering the full hiring cycle, from writing job descriptions to sending offer letters. It is built for recruiting teams and hiring managers who use ChatGPT, Claude, Copilot, or Gemini and want consistent, professional output without starting from a blank page every time.

What is in this pack

Twenty-plus prompts organized into six categories: job descriptions, candidate outreach, interview preparation, candidate evaluation, offer and rejection communications, and internal reporting. Each prompt is copyable, includes labeled variables so you can swap in role and company details, and works across the four major AI platforms without modification.

The pack sits alongside prompts organized by professional role and shares structure with similar libraries for marketing teams and business operations.

One honest caveat before you start: AI drafts for candidate-facing communication need a human read before they go out. Models will produce confident, fluent text that occasionally misreads tone or omits a legally relevant detail. Budget ten minutes of review per batch, not per message.

The rest of this page shows the prompts, example outputs, and guidance on adapting the pack for your team.

The prompts, by category

These prompts are grouped by the stage of the hiring workflow where they're most useful. Copy any prompt into ChatGPT, Claude, Copilot, or Gemini. Replace the bracketed variables with your own details before you run it.

Job descriptions

Write a job description from a hiring manager's notes

Variables: [ROLE TITLE], [TEAM NAME], [KEY RESPONSIBILITIES], [MUST-HAVE SKILLS], [NICE-TO-HAVE SKILLS], [LOCATION/REMOTE POLICY], [SALARY RANGE OR "competitive"]

You are an experienced recruiter writing a job description for a [ROLE TITLE] on the [TEAM NAME] team.

Key responsibilities:
[KEY RESPONSIBILITIES]

Must-have skills and experience:
[MUST-HAVE SKILLS]

Nice to have:
[NICE-TO-HAVE SKILLS]

Location and remote policy: [LOCATION/REMOTE POLICY]
Salary: [SALARY RANGE OR "competitive"]

Write a clear, honest job description in plain language. Avoid corporate buzzwords. Lead with what the person will actually do, not with company history. Keep it under 400 words. End with a two-sentence summary of who this role is right for.

When to use it: At the start of any new requisition, especially when the hiring manager's brief is vague or verbal. What to change first: Swap [KEY RESPONSIBILITIES] for bullet points directly from the intake call.

[NEEDS REAL OUTPUT]

Rewrite a bloated job description to cut it by 30%

Variables: [EXISTING JOB DESCRIPTION TEXT]

Here is a job description that is too long:

[EXISTING JOB DESCRIPTION TEXT]

Rewrite it so it is at least 30% shorter. Keep every genuine requirement. Cut anything that describes the company culture in vague terms ("fast-paced," "dynamic," "passionate team"). Do not add requirements that were not in the original.

When to use it: When you inherit a JD from a previous hire cycle or from a template that's grown out of hand. What to change first: If the original is already concise, adjust the target from 30% to 15%.

Write an inclusive job description for a technical role

Variables: [ROLE TITLE], [REQUIRED SKILLS], [YEARS EXPERIENCE], [TEAM CONTEXT]

Write a job description for a [ROLE TITLE] role that will attract a diverse candidate pool.

Required skills: [REQUIRED SKILLS]
Experience level: [YEARS EXPERIENCE]
Team context: [TEAM CONTEXT]

Rules:
- Replace any years-of-experience requirement with a skills-based equivalent where possible
- Avoid gendered language (do not use "rockstar," "ninja," "dominant")
- List only requirements that are genuinely required, not aspirational
- Use plain language throughout

When to use it: Any technical or senior hire where your applicant pool has been narrow in previous rounds. What to change first: The [REQUIRED SKILLS] list. Be honest about what's truly required versus what you'd like.

Candidate outreach

Write a cold outreach message for a passive candidate

Variables: [CANDIDATE NAME], [THEIR CURRENT ROLE], [THEIR NOTABLE WORK OR SIGNAL], [YOUR ROLE TITLE], [ONE SPECIFIC REASON THIS ROLE FITS THEM], [YOUR NAME]

Write a short LinkedIn outreach message to [CANDIDATE NAME], who currently works as [THEIR CURRENT ROLE].

Reason for reaching out: [THEIR NOTABLE WORK OR SIGNAL]
Role you're hiring for: [YOUR ROLE TITLE]
Why this role fits them specifically: [ONE SPECIFIC REASON THIS ROLE FITS THEM]
Sender: [YOUR NAME]

Keep it under 75 words. Sound like a person, not a recruiter template. Do not open with "I hope this message finds you well." End with a single low-friction question, not a request to apply.

When to use it: Sourcing outreach on LinkedIn or email for roles where inbound applicants haven't matched. What to change first: [ONE SPECIFIC REASON THIS ROLE FITS THEM]. Generic here makes the whole message worthless.

[NEEDS REAL OUTPUT]

Write a follow-up message after a candidate goes quiet

Variables: [CANDIDATE NAME], [ROLE TITLE], [LAST CONTACT DATE], [ONE NEW PIECE OF INFORMATION ABOUT THE ROLE]

Write a brief follow-up message to [CANDIDATE NAME] who expressed interest in the [ROLE TITLE] role. We last spoke on [LAST CONTACT DATE] and they have not responded since.

New information to include: [ONE NEW PIECE OF INFORMATION ABOUT THE ROLE]

Keep it under 60 words. Do not guilt them. Do not ask why they went quiet. Give them one new reason to re-engage and make it easy to say no if they're not interested.

When to use it: Any time a strong candidate stops responding after an initial positive exchange. What to change first: Make sure [ONE NEW PIECE OF INFORMATION] is genuinely new. A timeline update, a team change, or a salary revision all work.

Write a rejection message that leaves the door open

Variables: [CANDIDATE NAME], [ROLE TITLE], [SPECIFIC GENUINE POSITIVE FROM THEIR PROFILE OR CONVERSATION]

Write a rejection email for [CANDIDATE NAME] who applied for [ROLE TITLE].

Something genuine and specific to say about their profile: [SPECIFIC GENUINE POSITIVE FROM THEIR PROFILE OR CONVERSATION]

Rules:
- Do not use "we went with a candidate whose skills more closely aligned"
- Do not promise to keep them on file if you won't
- Be warm but brief (under 100 words)
- Leave the door genuinely open only if that's true. If it isn't, close cleanly.

When to use it: Post-interview rejections for candidates you'd consider again, or where a poor experience could hurt your employer brand. What to change first: If you genuinely won't reconsider them, remove the open-door language entirely and just close warmly.

Interview preparation

Build a structured interview scorecard

Variables: [ROLE TITLE], [COMPETENCIES TO ASSESS], [NUMBER OF INTERVIEWERS]

Create a structured interview scorecard for a [ROLE TITLE] role.

Competencies to assess: [COMPETENCIES TO ASSESS]
Number of interviewers: [NUMBER OF INTERVIEWERS]

For each competency, provide:
- One behavioral question
- One follow-up probe
- A 1-to-4 scoring rubric (1 = no evidence, 4 = strong evidence)

Format it as a table. Keep the language plain so any interviewer can use it without training.

When to use it: Before a new role's first interview round, especially when multiple interviewers need to align. What to change first: [COMPETENCIES TO ASSESS]. Pull these from the job description, not from a generic list.

[NEEDS REAL OUTPUT]

Write role-specific behavioral interview questions

Variables: [ROLE TITLE], [TOP 3 CHALLENGES THIS PERSON WILL FACE IN THEIR FIRST 6 MONTHS]

Write 8 behavioral interview questions for a [ROLE TITLE] role.

The biggest challenges in this role's first 6 months:
[TOP 3 CHALLENGES THIS PERSON WILL FACE IN THEIR FIRST 6 MONTHS]

Each question should:
- Follow the STAR format (Situation, Task, Action, Result)
- Be specific to the challenges listed above, not generic
- Avoid hypotheticals ("What would you do if...") in favor of past experience ("Tell me about a time...")

When to use it: Whenever you're opening a role that's new to your team or that has had consistent hiring mistakes in the past. What to change first: The [TOP 3 CHALLENGES]. Interview questions written against the wrong challenges are almost useless.

Draft a candidate briefing document for hiring managers

Variables: [CANDIDATE NAME], [ROLE TITLE], [CANDIDATE SUMMARY FROM CV OR LINKEDIN], [POINTS TO PROBE IN THE INTERVIEW]

Write a one-page briefing document for a hiring manager who is about to interview [CANDIDATE NAME] for the [ROLE TITLE] role.

Candidate background:
[CANDIDATE SUMMARY FROM CV OR LINKEDIN]

Areas to probe:
[POINTS TO PROBE IN THE INTERVIEW]

Include: a three-sentence candidate summary, the top two strengths evident from their background, the top two gaps or unknowns, and three suggested questions tailored to this specific candidate.

When to use it: Before any panel or hiring manager interview where the interviewer has limited prep time. What to change first: [POINTS TO PROBE]. This is where your recruiter judgment matters most. Don't skip it.

Screening and assessment

Write a take-home screening task brief

Variables: [ROLE TITLE], [CORE SKILL TO ASSESS], [TIME LIMIT], [FORMAT FOR SUBMISSION]

Write a take-home task brief for candidates applying for a [ROLE TITLE] role.

Core skill to assess: [CORE SKILL TO ASSESS]
Time limit: [TIME LIMIT]
Submission format: [FORMAT FOR SUBMISSION]

The brief should:
- Explain the purpose of the task honestly (candidates can tell when it's fake)
- Be completable within the stated time limit
- Include clear evaluation criteria so candidates know what good looks like
- Not require proprietary knowledge of your company to complete

When to use it: Any mid-to-senior hire where a CV alone doesn't give you enough signal on core skills. What to change first: [TIME LIMIT]. Be realistic. Asking for more than two hours of unpaid work will lose strong candidates who have options.

Write a phone screen script

Variables: [ROLE TITLE], [MUST-VERIFY REQUIREMENTS], [CULTURE OR TEAM FIT SIGNALS YOU NEED], [LENGTH OF SCREEN IN MINUTES]

Write a phone screen script for a [ROLE TITLE] role. The screen will last [LENGTH OF SCREEN IN MINUTES] minutes.

Non-negotiable requirements to verify: [MUST-VERIFY REQUIREMENTS]
Team or culture signals to listen for: [CULTURE OR TEAM FIT SIGNALS YOU NEED]

Structure:
1. Two-minute intro and agenda
2. Quick verification of must-haves (yes/no questions where possible)
3. Two open questions to assess fit signals
4. Candidate questions (leave 3 minutes)
5. Next steps

Write the full script including suggested spoken phrases, not just headers.

When to use it: High-volume roles where you need to screen consistently across a team of recruiters. What to change first: The must-have verification questions. These should be binary and fast.

[NEEDS REAL OUTPUT]

Summarize a candidate's CV for a hiring manager

Variables: [CV TEXT OR PASTED PROFILE], [ROLE TITLE], [TOP 3 THINGS THE HIRING MANAGER CARES ABOUT]

Summarize the following candidate profile for a hiring manager reviewing applications for a [ROLE TITLE] role.

What the hiring manager cares about most:
[TOP 3 THINGS THE HIRING MANAGER CARES ABOUT]

Candidate profile:
[CV TEXT OR PASTED PROFILE]

Write a summary of 150 words or fewer. Cover: years of relevant experience, the two or three things most relevant to this role, and any obvious gaps or questions. Do not editorialize. Do not describe the candidate as "passionate" or "driven."

When to use it: When forwarding candidates to a time-poor hiring manager who won't read a full CV. What to change first: The [TOP 3 THINGS]. Align these with the hiring manager in the intake call, not after.

Offer and close

Write an offer letter covering email

Variables: [CANDIDATE NAME], [ROLE TITLE], [START DATE], [SALARY], [KEY BENEFITS TO HIGHLIGHT], [DEADLINE FOR RESPONSE]

Write a warm, professional email to accompany a formal offer letter for [CANDIDATE NAME], who has been offered the [ROLE TITLE] role.

Details:
- Start date: [START DATE]
- Salary: [SALARY]
- Key benefits to highlight: [KEY BENEFITS TO HIGHLIGHT]
- Response deadline: [DEADLINE FOR RESPONSE]

Tone: Genuine and welcoming. This person has just made a big decision. Don't sound like legal boilerplate. Keep it under 200 words. End by telling them who to call if they have questions.

When to use it: Every offer. A cold email attached to a PDF sets the wrong tone at a high-stakes moment. What to change first: [KEY BENEFITS TO HIGHLIGHT]. Pick one or two that genuinely matter to this specific candidate based on what came up in interviews.

Write a counter-offer response script

Variables: [CANDIDATE NAME], [THEIR COUNTER REQUEST], [WHAT YOU CAN AND CANNOT FLEX ON]

Write a response to a counter-offer request from [CANDIDATE NAME].

Their request: [THEIR COUNTER REQUEST]
What you can flex on: [WHAT YOU CAN AND CANNOT FLEX ON]

Write two versions:
1. If you can meet their counter fully
2. If you can meet it partially or not at all

Both versions should be honest, not defensive. Do not imply the original offer was generous. If you cannot move, say so plainly and focus on the non-financial value of the role.

When to use it: When a strong candidate comes back with a counter before signing. What to change first: Be honest in [WHAT YOU CAN AND CANNOT FLEX ON] before you run this. The model can only work with what you give it.

Write a candidate withdrawal email

Variables: [CANDIDATE NAME], [ROLE TITLE], [YOUR NAME]

Write an email from [YOUR NAME] to [CANDIDATE NAME] withdrawing their candidacy for the [ROLE TITLE] role.

Reason: the role has been put on hold / filled internally / cancelled (delete as applicable).

Keep it under 80 words. Be direct about the reason. Do not apologize more than once. Thank them for their time without being hollow about it. Leave the relationship intact.

When to use it: When headcount is frozen mid-process or a role closes before an offer is made. What to change first: Pick one reason from the list and delete the others. Vague withdrawal messages damage trust.

[NEEDS REAL OUTPUT]

Employer brand and internal comms

Write a job ad for a social post

Variables: [ROLE TITLE], [TEAM NAME], [ONE GENUINELY INTERESTING THING ABOUT THE ROLE], [LINK TO JD], [PLATFORM: LinkedIn / X / Instagram]

Write a social post advertising the [ROLE TITLE] role on the [TEAM NAME] team.

Platform: [PLATFORM]
One genuinely interesting detail about this role: [ONE GENUINELY INTERESTING THING ABOUT THE ROLE]
Link: [LINK TO JD]

Do not use the phrase "we're hiring." Do not open with an emoji. Lead with the interesting detail. Keep it under 150 words for LinkedIn, 280 characters for X. Sound like a person, not a brand account.

When to use it: When you want organic reach beyond your careers page, particularly for specialist or senior roles. What to change first: [ONE GENUINELY INTERESTING THING]. If you can't fill that in, the post will not work regardless of the prompt.

Write an internal referral request message

Variables: [ROLE TITLE], [WHAT MAKES THIS ROLE DISTINCTIVE], [REFERRAL BONUS IF APPLICABLE], [HOW TO REFER]

Write a short internal message asking employees to refer candidates for the [ROLE TITLE] role.

What makes this role genuinely interesting or unusual: [WHAT MAKES THIS ROLE DISTINCTIVE]
Referral bonus: [REFERRAL BONUS IF APPLICABLE]
How to refer: [HOW TO REFER]

Rules:
- Do not use "amazing opportunity"
-

## Real example outputs

The five prompts marked [NEEDS REAL OUTPUT] below will be updated with captured model responses once the pack is tested against each platform. What to look for in each case is noted alongside the placeholder.

| Prompt | Platform tested | What to notice |
|---|---|---|
| Job ad from a job description | ChatGPT | Whether the output drops jargon and leads with candidate benefit |
| Candidate outreach message | Claude | Whether the tone stays warm without sounding generic |
| Interview question set | Gemini | Whether questions map cleanly to the competencies listed |
| Rejection email | Copilot | Whether the message stays respectful without being vague |
| Hiring manager kickoff questions | ChatGPT | Whether the questions surface scope and success criteria, not just duties |

[NEEDS REAL OUTPUT x5]

:::keypoint heading="Worth pausing on"
Model outputs vary between platforms and change over time. What Claude produces today may differ from what it produces next month. Treat any captured output as a benchmark, not a guarantee, and re-test prompts whenever results start to drift.
:::

## How to adapt these for your team

The prompts above work out of the box for most roles, but a few small changes will make them significantly more reliable across your team.

Start with your company's standard language. If your job postings always use "hybrid" rather than "remote-first", or your rejection letters open with a specific phrase, add that as a standing instruction at the top of any prompt you reuse. Most AI tools let you save a system prompt or custom instruction for exactly this purpose.

Next, lock down anything that should not vary: seniority language, legal disclaimers on job ads, interview scoring rubrics. Treat those as fixed text that the model fills around rather than rewrites.

:::keypoint heading="Worth pausing on"
The prompts that go wrong fastest are the ones with too many variables left open. Specify the tone, the audience, and the output length, and you will get consistent results across different team members using different AI tools.
:::

For teams running high-volume hiring, a [shared folder](/product/shared-folders) is the most practical way to keep everyone using the same vetted versions rather than writing from scratch each time.

## Save this folder to your workspace

Running these prompts from a shared notes doc or a browser bookmark works fine for solo use. When your team grows, or when a second recruiter joins a search, that approach breaks down fast. Someone rewrites a prompt slightly, the original gets lost, and two people are sending candidates through different versions of the same outreach.

[Convergence shared folders](/product/shared-folders) let you store the full set in one place, assign prompts to specific roles or requisitions, and push updates to everyone at once. If you refine the screening-question prompt for a technical hire, the whole team picks up the change without a Slack message or a forwarded doc.

:::cta heading="Want your recruiting team working from the same prompts?"
Store, organize, and share this full prompt pack in a Convergence workspace. Works across ChatGPT, Claude, Copilot, and Gemini.
:::

This pack sits alongside [prompts for other business functions](/prompts/business) in the broader Convergence library, so when recruiting overlaps with HR, legal review, or marketing (employer brand copy, for instance), the relevant prompts are one folder away.

## Frequently asked questions

### Do these prompts work across ChatGPT, Claude, Copilot, and Gemini?

Yes. Every prompt in this pack is written in plain language and works across all four major AI tools without modification. The bracketed variables are the only things you need to change before running them.

### Can AI write a full job description without any input from me?

No, and you should not want it to. A prompt with no role context produces generic output that looks like every other posting. The prompts here are designed to take your specific inputs, seniority level, team context, compensation range, and produce something worth publishing.

### Is AI reliable for screening decisions?

No. These prompts help you organize and analyze information faster, but final screening and selection decisions should always involve a human. Using AI output as a sole basis for rejecting candidates raises real legal and ethical concerns.

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

[Convergence shared folders](/product/shared-folders) let your team access, run, and update prompts from one place, across whichever AI tools they already use.

### Where can I find prompts for other roles?

The [full prompts-by-role library](/prompts) covers functions beyond recruiting, including [prompts for marketers](/prompts/marketing) and [prompts for business teams](/prompts/business).