convez.ai
SkillProdutividadeIntermediário
brianjkopp-netizen

ATS Score Builder

brianjkopp-netizen/ats-score-builder

Otimize seu currículo para passar em filtros ATS sem inventar qualificações

Ver no GitHub
InicialPluginScore · de 0 a 100

Skill reutilizável que otimiza currículos para ATS (Applicant Tracking Systems) atingindo 80%+ de compatibilidade de palavras-chave, sem fabricar experiência. Usa um checklist de 15 passos e padrões de formatação para melhorar parseabilidade e correspondência exata de termos com a descrição da vaga.

Instalação

$npx convez-ai brianjkopp-netizen/ats-score-builder

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
0
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 5 meses
Versão
—
Issues abertas
0
Contribuidores
1

Casos de uso

  • Otimizar currículo para vaga específica
  • Aumentar taxa de aprovação em ATS
  • Aplicar para posições acima do nível atual

Veredito

Skill prática e honesta que endereça um problema real: filtros ATS rejeitam candidatos qualificados. O diferencial é a recusa em fabricar experiência, focando apenas em reempacotar qualificações reais. Útil para quem busca melhorar taxa de sucesso em processos de seleção.

README

A tradução para o português deste README ainda está na fila.

ATS Score Builder

A reusable, portable skill that tailors any resume to any job description to clear an 80%+ ATS match (measured by any ATS keyword-match checker). It works as a Claude Skill, and the same checklist works on its own as a manual process with any AI assistant or by hand.

It encodes a method refined across dozens of real applications: one that took resumes from the 50s into the 90s on ATS keyword scans, and turned ATS rejections into recruiter screens.


What it is

Applicant Tracking Systems (ATS) reject most resumes before a human ever reads them, usually for reasons that have nothing to do with the candidate's qualifications. They reject on two things:

  1. Parseability — whether the software can cleanly read your file (tables, graphics, odd bullets, and non-standard dates quietly break this).
  2. Keyword match — whether your resume contains the job description's exact phrases, scored as literal string matching. "Agile" does not satisfy "Agile software development." "AI" does not satisfy "Artificial Intelligence."

ATS Score Builder is a 15-step checklist plus formatting standards that fix both. It does not write a resume for you from nothing, and it never invents experience. It surfaces the real qualifications you already have in the exact language the ATS is scoring.

What it does

  • Extracts the keywords a specific job description scores hardest (the ones appearing 2+ times, plus implicit skills the responsibilities imply).
  • Flags the gap between what your resume says and what the JD wants, in exact-match terms.
  • Enforces the acronym rule: every key term appears in both its acronym and spelled-out form.
  • Weaves required soft skills into your summary instead of burying them in a list.
  • Applies a high-impact ALL-CAPS bullet-label technique that raises keyword density while making the resume easier for a human recruiter to scan.
  • Fixes the silent formatting killers: bold overuse, non-standard bullets, "&" and pipe characters, long-form dates, and over-long paragraphs.
  • Includes scope-gap framing for applying one level above your current title.
  • Gates on a sign-off step so you never fabricate a qualification you do not hold.

What it deliberately does not do

It will not invent experience, skills, titles, or metrics. An honest 80%+ resume gets a real candidate past the filter and in front of a human, which is the only thing the score is for. A resume gamed with skills you do not have wastes everyone's time at the interview.


What's in this package

ats-score-builder/
├── SKILL.md                          # The skill itself: the 15-step checklist + formatting standards
├── README.md                         # This file
├── LICENSE                           # MIT
├── templates/
│   └── resume_template.md            # Blank ATS-safe resume structure to build from
└── examples/
    ├── sample_job_description.md     # A realistic JD to practice against
    ├── before_resume.md              # A strong-but-untailored resume (scores low)
    ├── keyword_audit.md              # The worked audit connecting before to after
    └── after_resume.md               # The same resume after the skill (scores high)

Read the three example files in order — before, audit, after — to see exactly what the method changes and why. Nothing in the "after" resume is fabricated; every change is wording or formatting.


How to use it

Option A — As a Claude Skill (recommended)

Claude Skills are folders containing a SKILL.md with YAML frontmatter. To use this one:

  1. Download this repository (Code → Download ZIP, or git clone).
  2. Install the skill wherever your Claude environment loads skills from. In Claude Code or the Claude apps that support skills, place the ats-score-builder/ folder in your skills directory. Claude reads the name and description in SKILL.md and triggers the skill automatically when you ask it to tailor or score a resume.
  3. Use it. Paste a job description and your current resume and say something like "Tailor my resume to this job and get me past the ATS." The skill runs the keyword audit first, shows you matched and unmatched keywords, waits for your sign-off, then builds the tailored version.

If your environment does not auto-load skills, you can paste the contents of SKILL.md into the conversation as instructions and it works the same way.

Option B — As a manual checklist (no special tooling)

The method is tool-agnostic. Open SKILL.md, work through the 15 steps against your resume and a specific job posting, and apply the formatting standards. Use templates/resume_template.md as your starting structure and the examples/ files as your reference.

Verifying the score

Run the finished resume through an ATS keyword-match checker (several free and paid options exist; DOCX scans more reliably than PDF). Record any missed Hard Skills and Soft Skills, add those exact strings, and rescan.

  • Target: 80%+ minimum, 90%+ for high-priority roles.
  • A near-100% score is achievable but optional; for senior roles, the high 80s to low 90s can read more naturally to an experienced reviewer. That is a judgment call, not a rule.

The 15 steps at a glance

  1. Extract JD phrases appearing 2+ times, plus implicit skills; flag proprietary terms to exclude.
  2. Verify exact-match wording, not synonyms.
  3. Include both acronym and spelled-out forms of every key term.
  4. Weave required soft skills into the summary prose.
  5. Put the exact target job title in the tagline.
  6. Put city and state in the contact line.
  7. Keep paragraphs under ~40 words.
  8. Bold only company names and job titles.
  9. Use standard round bullets only.
  10. Replace "&" with "and"; keep pipes out of the body.
  11. Use consistent MM/YYYY dates.
  12. Lead each bullet with an ALL-CAPS exact-JD-phrase label.
  13. Frame scope when applying above your current title.
  14. Sign off on the keyword list before writing — never fabricate.
  15. Scan, record red-X items, iterate to the target score.

Full detail and rationale for each step is in SKILL.md.


Contributing

Issues and pull requests welcome, especially additional worked before/after examples in different fields (the bundled example is product management; nursing, software engineering, finance, and trades examples would all be useful).

License

MIT — see LICENSE. Use it, fork it, adapt it. Attribution appreciated but not required.

A note on intent

This method helps qualified people get read. It is for getting your real experience past an automated filter, not for gaming your way into a role you are not equipped for. Used honestly, it levels a playing field that formatting and keyword quirks have tilted against good candidates.

Ver o original no GitHub

Bom para

Relacionados