Can the software actually read your file? A specific answer.
Most of this category sells "pass the ATS", which is not a real thing. An applicant tracking system is software each employer configures with their own variables; there can be several screens for one job, and nobody can promise you get through them.
What can be answered is narrower and more useful: can a text parser read this file at all? Are your name, email, phone, employers, titles and dates extractable? Is anything trapped in a table, a column, a header or an image? Those are facts about a document, and we check twelve of them.

The one you have actually been sending. No account, no card, nothing to sign up for.
The twelve readability checks are deterministic — no AI, no judgement calls. They run instantly and we can defend every one of them line by line.
Each check names the problem in plain language. The ones a rebuild fixes automatically are marked; the ones we cannot invent for you — a missing phone number, say — are left alone and flagged as yours to fix.
Continue into the builder and the fixable problems are gone in the rebuilt version. We already parsed your file, so it costs you no second wait.
These are constraints we chose, not gaps we haven't got to.
There is no score out of 100 here and there never will be. A number teaches you to chase the number, and it implies a precision about a hiring manager that nobody has.
This half is about the file. How well your material matches one specific posting is a different question, needs the posting as an input, and gets answered in bands — see Job Sourcing and Targeted Applications.