cv.shykov.dev
ATS Resume Toolkit
100% in-browser resume analysis

Free CV ATS score checker. Private by default.

Get your CV ATS score from a PDF or DOCX in seconds, see the highest-impact fixes first, match keywords, and build an ATS-clean resume without uploads, accounts, or LLM calls.

Runs in your browserNo uploadsNo LLMPDF & DOCXOpen source
In your browser
Score, top fixes, keyword signals
83
Parseability
Sections
Content
1

Choose PDF/DOCX

Pick a resume file from your device.

2

Parse in your browser

Text extraction and scoring run in this browser.

3

Fix the top issues

See score, breakdown, keywords, and exportable notes.

No server uploadDeterministic checks, no LLM callsOpen source, MIT license

Loading fast ATS score…

Privacy claims you can inspect.

The code is open source under the MIT license, and the product promise stays narrow: local document parsing, deterministic scoring, no account wall, and no AI model call hidden behind the interface.

Browser-side parser

PDF and DOCX text is extracted in the browser with client-side libraries.

No LLM scoring

Scores come from repeatable checks for parseability, sections, format, and content signals.

Open source on GitHub

MIT-licensed. Read the scoring checks, parser, and export flow, open an issue, or send a pull request.

What is an ATS score?

A fast heuristic check for parser-friendly text, contact details, sections, dates, bullets, and measurable impact.

Is my CV uploaded?

No. PDF and DOCX files are processed locally in your browser and are not sent to a server.

Does this use AI?

No. The score is deterministic and does not use LLMs, model calls, accounts, or API keys.

PDF or DOCX?

Both work. Text-based PDFs give the clearest page-count and parseability signal.

How accurate is the score?

It estimates how well a parser will read your document. It is a heuristic, not a prediction of a recruiter’s decision or of any employer’s system, and it has not been validated against hiring outcomes. Other checkers give the same file different numbers because each one measures something different.

Can the score be biased?

There is no model and no training data, so there is no dataset to be biased. The rules still carry assumptions, such as the section headings and CV conventions they accept, and the writing-style check covers English CVs only and can flag good writing. It is not an AI detector. Nothing is decided about you: each point shows the rule behind it, and you decide what to change.

Guides