resume-tailoring
Skill0 starsChanged 2 months agoMITView on GitHub
Tailor a master resume to one specific job posting: keyword extraction, language matching, two-page enforcement, and the bullet-rhythm pass that keeps it from reading machine-written. Use whenever building or revising a resume for a named role.
7 other projects use a version of this.
Adopted by
Most recently taken up by
- lengrufeng2008/dsh-skill-librarytheir own version, first seen Nov 15, 2026
- eigent-ai/agent-skillstheir own version, first seen Nov 1, 2026
- amit-t/skillstheir own version, first seen Oct 25, 2026
- jaron-goh-centr/claude-skillstheir own version, first seen Oct 11, 2026
- miteshjadav98/ClaudeSkillstheir own version, first seen Aug 7, 2026
- amanattar/resume-tailoring-skilltheir own version, first seen Mar 18, 2026
Other versions7 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 7
- varunr89/resume-tailoring-skill★ 750same name
- amanattar/resume-tailoring-skill★ 116same name
- eigent-ai/agent-skills★ 27same name
- amit-t/skills★ 4same name
- jaron-goh-centr/claude-skills★ 0same name
- lengrufeng2008/dsh-skill-library★ 0same name
- miteshjadav98/ClaudeSkills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository3 other files it gives its agents.
- humanizeSkill
- job-search-writingSkill
- social-trends-radarSkill
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 8 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 7 edited versions we read keep these.
8
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
No keyword stuffing: If a term does not describe work you actually did, leave it out.
- Distinctive: line 33
No employment gaps: Keep the older short roles that bridge them.
- Distinctive: line 37
Never set type smaller than 10pt.
- Distinctive: line 47
Do this on every build, not only when something feels off.
- Distinctive: line 75
No type under 10pt anywhere.
- Distinctive: line 77
No three-plus-word phrase used more than twice.
- Distinctive: line 78
Every keyword claim maps to something you can talk about for two minutes.
- Distinctive: line 79
Read five consecutive bullets aloud.
“Common” is measured: kept by at least 6 in 10 of the 7 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Resume tailoring
Build a version of the resume aimed at one posting. Start from the master resume every time and cut toward the role, rather than writing fresh and hoping the good material survives.
Read the posting first
Pull the top keywords out of the job description before touching the resume. Use each one once or twice in the final document. That is the whole budget.
Then check the language you already have against the language the posting uses. Where they mean the same thing, take theirs. "Directed" becomes governed, led, managed, or spearheaded depending on which word the posting itself reaches for. The point is not to flatter the ATS, it is that a hiring manager skimming for their own vocabulary finds it.
Two limits on that:
No exact phrase of three or more words appears more than twice in one resume. Repeat a long phrase a third time and the document starts to read as written for a parser instead of a person.
No keyword stuffing. If a term does not describe work you actually did, leave it out. A resume that clears the filter and then collapses in the screening call is worse than one that never cleared it.
Structure
Two pages maximum. This is a hard cap, not a target.
- Summary opens with a thesis about level, not a recap of history. Audience scale, channels owned, crisis readiness, executive exposure. The reader should place your seniority in the first four or five lines.
- Bullets lead with outcomes and carry metrics where they exist. Growth percentages, reach, campaign results. Responsibilities described without results read as a job description rather than a record.
- Core skills in a single column. Two columns parse unpredictably and save less room than they appear to.
- Tools and Education stay separate sections, never merged.
- Dates on the title row, formatted
Title | Mon YYYY - Mon YYYY, not on the company row. Cleaner for ATS parsing and easier for a human to scan. - One degree per line, degree and school spelled out in full.
- Labels are bold wherever they appear, including labels that share a line with other content.
- No employment gaps. Keep the older short roles that bridge them.
Formatting floors
Never set type smaller than 10pt. When a resume runs long, trim bullets. Do not shrink the font, and do not narrow the margins past what is already tight. Cutting a weak bullet costs nothing; unreadable type costs the interview.
Bold the numbers in results so a skimming eye catches them.
At most one em dash in the entire document, and zero is better. Commas, colons, and parentheses do the same work without the tell.
Make the contact line real hyperlinks (phone, email, site, LinkedIn), underlined in the same color as the surrounding text rather than default link blue.
The rhythm pass
Do this on every build, not only when something feels off.
A resume can be entirely true, well formatted, and still read as machine-written, because the thing that gives it away is structure rather than vocabulary. One real example: a finished draft scored 75% AI-generated by a detector. Nothing in it was invented. Every bullet ran the identical pattern, verb plus participial phrase plus participial phrase plus metric. Three-item lists throughout. Every sentence about the same length.
The fix:
- Break the uniform bullet shape. Mix one-clause bullets with two-sentence bullets. Let a metric have its own short sentence. "Instagram grew 7% in three months."
- Cut participial tails that interpret instead of report. "...protecting craft and visual cohesion across every brand surface" is filler wearing a suit. "Craft is the part I won't compromise on" says something.
- Break the rule of three. Two items, or four. Not three in every bullet.
- Use first person sparingly. Three or four instances across the page is a strong human signal. In every bullet it becomes its own pattern.
- Give the summary one short punchy sentence against the longer ones. "Reach is the easy part."
Facts never change in this pass. Numbers, dates, titles, keywords all stay exactly as they were. You are changing rhythm, not content.
A word on the detectors themselves: they are unreliable on resumes, because dense metrics and parallel structure look machine-written even when a person wrote every word. Treat a score as a prompt to look, not as a verdict. The patterns above are worth removing regardless of what any tool says about them.
Phrases to avoid outright
Known AI resume language, and transparently so to anyone who reads resumes for a living:
- "at the intersection of"
- "demonstrating"
- "orchestrated"
- "helmed"
Before delivering
- Page count verified by rendering, not estimated.
- No type under 10pt anywhere.
- Em dashes: zero or one.
- No three-plus-word phrase used more than twice.
- Every keyword claim maps to something you can talk about for two minutes.
- Read five consecutive bullets aloud. If they have the same shape, go back to the rhythm pass.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

