At a glance
A 2025 TopResume hiring-manager survey found 52% of hiring managers said AI-assisted writing in cover letters was acceptable. The dealbreaker was submitting unedited AI output. The seven prompts below are designed to produce cover letters worth editing into your own voice, not finished drafts to copy-paste. The workflow goes: research the company, generate three angles, pick the strongest, draft the letter, edit the AI tells, send. Each step has a prompt.
Why does the same cover letter that wins one offer get rejected by another?
The average cover letter reviewer in 2026 spends 7 to 11 seconds on the page. Ten years ago it was closer to half a minute. That collapse in attention happened at the same time roughly 60-65% of large employers added automated ATS screening with AI-detection signals (estimate from recruiting industry surveys). The job seeker pasting Claude’s first draft straight into the form is now getting filtered out by both layers at once: the ATS flags the AI patterns, and the human, if the letter survives the filter, reads three sentences and moves on.
The solution is not to skip AI. The TopResume survey is clear that hiring managers do not mind AI assistance. They mind unedited AI output. The workflow below uses Claude as a research, drafting, and editing partner without ever sending what it writes directly. Every prompt is a step in the workflow. Each one has constraints baked in that prevent the most common AI tells from appearing in the output.
⚠️ What gets a cover letter flagged in 2026
Em-dashes used in every paragraph. “I am writing to express my interest in” openers. “I would welcome the opportunity to” closers. Triple-adjective phrases (“dynamic, results-driven, passionate”). The words “comprehensive,” “robust,” “synergize,” “leverage,” “passionate about.” Closing rhetorical questions. The “I noticed your recent…” opener used without naming the specific recent thing. Our full guide on what AI detectors and human recruiters look for: How to Edit AI Out of Your Writing.
What is the cover-letter workflow, step by step?
Seven prompts, run in sequence. Roughly 20 minutes of actual work per cover letter, including the research. The output of one prompt feeds the next. Run them out of order and the letter will sound generic, because each prompt is designed to add a specific layer of specificity that the previous prompt does not have.
The 7-step workflow
- Research, find three real things about the company beyond the about page
- Generate angles, draft three different framings of why you fit
- Pick the strongest, choose the one with the most specific connection
- Draft, write a 250-word letter around that angle
- Tighten, cut the AI tells and reduce to 180-220 words
- Personalize, add the one sentence only a real human reader could write
- Final check, run the pre-publish prompt on the result
Step 1: Research the company
Here is the job description: [paste it].
Here are the company links I have access to: [LinkedIn page / website / recent press / podcast / blog].
Help me find three specific things about this company that I could plausibly reference in a cover letter:
– One thing about their work in the past 90 days (product launch, funding, hiring push, customer win)
– One thing about their culture or values that you can verify from their materials, not their about page
– One thing about the team I would be working with (if information is public)
For each, tell me how to verify it before I cite it, and what to do if the source is just one tweet.
Why this works: Most cover letters reference the company’s mission statement or “about” page. Recruiters have read those words a thousand times. The “past 90 days, culture beyond the about page, the team” framework forces Claude to surface specifics the recruiter has not already heard. The “how to verify” output is what separates a researched letter from a hallucinated one.
Step 2: Generate three different angles
Here are the three things I learned about the company: [paste output of step 1].
Here is the job description: [paste].
Draft me three different angles for why I am a strong fit. Each angle should be one sentence.
– Angle 1: connect a specific experience on my resume to a specific company need
– Angle 2: connect something the company is doing right now to a skill I have
– Angle 3: be candid about what I do not yet have, and what I would bring that compensates
For each angle, tell me which kind of recruiter responds to it (early-career vs senior, IC vs manager, etc).
Why this works: Most people pick the first angle that comes to mind. Asking Claude for three forces you to compare. The third angle, the “be candid about what I do not yet have” one, is what veteran recruiters actually find compelling, because it reads as a human writing not a candidate performing. The recruiter-type tag helps you pick the right angle for the role.
Step 3: Pick the strongest angle
Here is the recruiter likely to read this: [name + title from LinkedIn, or “unknown, large enterprise”].
Help me pick the strongest of the three angles for this specific recruiter.
Output: the angle you would pick, the one-sentence reason, and what could go wrong with it if I am wrong about the recruiter.
Why this works: Most people skip this step and pick the angle that flatters them. Asking Claude to pick gives you a tie-breaker that is not your ego. The “what could go wrong” output is the part that makes you sharper on the rebound.
Step 4: Draft the letter
Length: 250 words. The recruiter will read 11 seconds, so the first sentence has to land.
Constraints:
– No “I am writing to express my interest in”
– No “I would welcome the opportunity to”
– No triple-adjective phrases (no “passionate, dynamic, results-driven”)
– No “leveraged” or “utilized” or “comprehensive”
– One em-dash maximum
– Use specific numbers from my resume where they fit
Output: the letter plus the one sentence I should personalize further before sending.
Why this works: Almost every “AI cover letter” failure mode is in this prompt’s negative-instruction list. Claude obeys explicit constraints when you spell them out. The “11 seconds” frame puts the burden on the first sentence, which is where most cover letters are weakest. The “personalize further” placeholder reminds you the AI cannot finish the job.
Step 5: Tighten to 180-220 words
Cut it to 180-220 words while keeping every concrete specific (company facts, my resume numbers, the angle).
For each sentence you cut, tell me one reason why.
For each sentence you kept, tell me what it earns.
Reject any sentence that could appear in another candidate’s letter unchanged. Those are the ones that scream AI.
Why this works: The “could appear in another candidate’s letter unchanged” test is the single best filter for generic AI prose. Asking Claude to be the editor with that rule gets you a tighter letter than asking it to draft a shorter one from scratch. The cut-and-keep reasoning teaches you what your letter is doing.
Step 6: Add the human-only sentence
Help me add one sentence that only I could write, in the second or third paragraph.
The sentence should reference something a real person who knows my work would know:
– A specific project I led (not the title, the messy detail)
– A specific decision I made and what I learned
– A specific phrase or framework I use that is unusual
– A specific failure I owned and how I rebounded
Pick one and draft three versions. Tell me which I should use based on the tone of the rest of the letter.
Why this works: This is the step that makes the difference between an AI-edited letter and a real letter. AI cannot write your fingerprint sentence because it does not know your actual stories. Asking it to give you three possible scaffolds, then choosing the one that fits, is how you put a piece of yourself into the letter without freezing on a blank page.
Step 7: Final pre-send check
Before I submit it, run this 5-point check:
1. Are there any em-dashes I should replace with commas, parentheses, or periods?
2. Are there any triple-adjective phrases I should cut to one strong adjective?
3. Does the first sentence earn the next sentence?
4. Is there a sentence that could appear in another candidate’s letter unchanged?
5. Did I name a specific recent thing about the company that I verified?
For each point, tell me what to fix or confirm I am clean.
Why this works: The 5-point check is the cover-letter version of the full pre-publish pass in our How to Edit AI Out of Your Writing guide. The questions are the most common things that get a cover letter rejected at the AI-detection layer or the 11-second human-read layer. Run them in 90 seconds before you hit submit.
What is the worst thing you can do with AI on a cover letter?
- Send the first draft. The TopResume survey is clear: unedited AI output is the dealbreaker. The hiring manager who said AI is fine in proofreading and drafting is the same one who rejects the candidate who sent ChatGPT’s first paragraph verbatim.
- Use the same letter for 50 applications. ATS systems can fingerprint near-identical letters across applications. So can the recruiters who run those ATS systems. Step 1 of the workflow is non-negotiable per company.
- Let AI invent specifics. Hallucinated company details (“I admire your recent expansion into Latin America”) catch eventually, especially in late-stage interviews when someone asks you to elaborate. Step 1’s “how to verify” output is the safeguard.
- Use the rhetorical-question closer. “I would love to discuss how my background could contribute to [company]’s mission. Looking forward to your reply.” Both sentences are the exact signature of AI prose. Cut them.
- Disclose AI use unprompted. Most hiring managers do not want you to confess. They want a good letter. If the application explicitly asks, disclose accurately. If it does not, just send a good letter.
What if you find yourself running this workflow every week?
If you are job-searching and writing five cover letters a week, the workflow above is a candidate for the next tier of the ladder. Save the seven prompts as Claude Code skills. Bundle them as a plugin called cover-letter-pipeline that you invoke with one command. Pass in the company name and job description; the pipeline runs all seven steps and hands you a final letter to edit. Setup time: an hour. Time saved per application: 15 minutes. After ten applications, the setup pays for itself.
📊 The Prompt-to-Workflow Ladder
Tier 1: the prompts (this post). Tier 2: the skill (saved as a file). Tier 3: the plugin (bundle of seven skills into one cover-letter pipeline). Tier 4: the workflow (the pipeline fires automatically when you bookmark a job posting). When to climb →
What are the common questions about AI cover letters?
Will the ATS reject my cover letter for being AI-written?
Not for being AI-assisted. Probably for being AI-pattern-matching. The seven-step workflow exists to strip the patterns. Run through it once and the resulting letter will not pattern-match.
How long should my cover letter be?
180-220 words. Anything longer assumes a reader who has time. Recruiters do not. Anything shorter looks like you did not bother.
Should I name the recruiter or use “Dear Hiring Manager”?
Name them if you can find them on LinkedIn or the company’s careers page. “Dear Hiring Manager” reads as AI default. If you cannot find a name, “Dear [Company] team” or “Dear [Department] hiring team” is better than the generic.
Should I tell the company I used AI?
Only if they ask. The TopResume survey found 52% of hiring managers accept AI-assisted writing. Their concern is unedited output, not the use itself. If the application form asks, answer accurately. If it does not, the silence is the standard.
Where do these prompts come from?
They are the seven most-used cover-letter prompts in the larger AI Prompt Library. The Library has over 500 prompts across 33+ categories with three difficulty levels, including a career section with prompts for resumes, LinkedIn updates, salary negotiation, and interview prep.
Sources to read next?
- TopResume career advice (the recruiting-industry surveys cited)
- Jobscan blog on ATS detection
- LinkedIn Talent Solutions guidance for recruiters
- SHRM (Society for Human Resource Management)
- Applicant Tracking System (Grokipedia)
✏️ Before you submit
Cover letters that pattern-match as AI get screened out by ATS and skimmed past by humans. Our pre-publish editing prompt and the manual five-pass checklist: How to Edit AI Out of Your Writing → The full 29-pattern catalog from Wikipedia's “Signs of AI writing” guide, plus the open-source Humanizer skill (Claude Code / OpenCode, MIT, 20K+ stars) that runs the list, is documented there.
The AI Prompt Library · $39
over 500 tested prompts including the career section
The seven cover-letter prompts above are a free preview. The full Library has prompts for resumes, LinkedIn updates, interview prep, salary negotiation, and the harder situations: career pivots, ageism, gaps, returning to work, layoffs.
1-on-1 Custom AI Tutorial with James · $99
Want help running the workflow on your actual job hunt?
A 1-hour private call. We run the seven-step workflow on a real job posting together, then build the cover-letter skill and resume-update skill so the next ten applications take 5 minutes each.
Get Smarter About AI Every Morning
Free daily newsletter. Built for people who want to use AI well, not chase every model.
Free forever. Unsubscribe anytime.
You may also like
- Best AI Prompts for Resume Writing — what gets you in the door
- Best AI Prompts for Job Interviews — the room you walk into
- Best Prompts for Salary Negotiation — the conversation that pays for itself
- How to Edit AI Out of Your Writing — strip AI patterns before you publish
- Prompt to Workflow: The AI Ladder — when to climb from prompt to workflow