How is a finding structured?

Every finding carries a status, a confidence, its evidence and a remediation link. Evidence means an observed value AND an expected value — an observed value alone is an assertion, not evidence. One finding is one (property, value) pair with a usage count, never one per element.

Why does the tool report criteria it cannot measure?

Because hiding them reads as not knowing they exist. All 56 criteria appear in the report with their name and status; the ones nothing automated can reach are shown and excluded from the estimate rather than quietly dropped or silently failed.

Will this match the Authority’s own assessment?

No, and we do not claim it will. The Authority assesses with a different methodology and its assessment is the reference. What this measurement is for is seeing the gap early, with evidence, while there is still time to close it.

Can an agency cite this to a client?

That is what the independence is for. The measurement is independent of both the agency and its client, every finding carries its evidence, and the methodology behind it is published rather than proprietary.

Methodology

We measure what can be measured, and say plainly what cannot

XFix is an independent tool that helps the teams behind Saudi government platforms see where they stand against the Platforms Code before the official review — as an honest estimate in the shape of a range, rather than a reassuring number that is not true.

How we work · the methodology in full

Nothing in the methodology is hidden. Every step is declared, and every criterion in the report states how it was measured and by which rule.

1A scan that does not burden the serverPublicly published pages only, one request per second, full compliance with robots.txt, and no form is ever submitted.

2Measured at three viewport widthsDesktop, tablet and mobile, because the responsive and accessibility criteria resolve differently at each.

3A declared rule for every criterion36 criteria have automated rules running today, 3 are judged by a language model at a lower confidence ceiling, 12 are automatable and have no rule built yet, and 5 can only be measured by a specialist.

4A range with two declared endsThe lower bound assumes the worst about anything not measured; the upper bound assumes the best. No single number is issued that would hide the distance between them.

Why a range and not a number?

39 of the 56 criteria issue a verdict today, 12 are automatable with no rule built yet, and 5 can only be measured by a specialist. Any tool that issues a single number from an automated scan is claiming to know what it did not measure. We show both ends: the lower assumes everything unverified fails, the upper assumes it passes, and the distance between them is itself information. What is measured, in detail.

A scan covers up to 8 pages. Beneath the 56 criteria sit 127 checkpoints, and the checkpoint is the operative unit — see the full index.

Who we are

XFix is a product of UX Arabia. The measurement rules behind it come from real Platforms Code compliance work with government entities, which is the half a reading of the document cannot produce.

Fully independent. No funding from and no affiliation with any government entity, and no entity can pay to improve its estimate.
Your data is yours. Scan results are not shared, not published, and not aggregated into public rankings.
The methodology is public. The measurement rules are published, and criticism of them reaches us and improves them.
Field experience, not a reading of the document. The measurement rules come from real compliance projects with government entities, not from the text of the Code alone.

What we are not

We are not an accreditation body. We issue no certificates, no seals and no “trust marks”.
We do not speak for the Digital Government Authority or interpret its intent. We measure the published text of the Code as it stands.
We do not promise that our estimate matches the outcome of the official review. The two methodologies are different.
We do not sell data, and we do not use your scans to train anything.

Frequently asked questions

Why a range instead of a score?: 39 of the 56 criteria issue a verdict today, 12 are automatable with no rule built yet, and 5 can only be measured by a specialist. Any tool that issues a single number from an automated scan is claiming to know what it did not measure. We show both ends instead: the lower assumes everything unverified fails, the upper assumes it passes, and the distance between them is itself information.
How is a finding structured?: Every finding carries a status, a confidence, its evidence and a remediation link. Evidence means an observed value AND an expected value — an observed value alone is an assertion, not evidence. One finding is one (property, value) pair with a usage count, never one per element.
Why does the tool report criteria it cannot measure?: Because hiding them reads as not knowing they exist. All 56 criteria appear in the report with their name and status; the ones nothing automated can reach are shown and excluded from the estimate rather than quietly dropped or silently failed.
Will this match the Authority’s own assessment?: No, and we do not claim it will. The Authority assesses with a different methodology and its assessment is the reference. What this measurement is for is seeing the gap early, with evidence, while there is still time to close it.
Can an agency cite this to a client?: That is what the independence is for. The measurement is independent of both the agency and its client, every finding carries its evidence, and the methodology behind it is published rather than proprietary.

The limits of this tool

XFix is an independent commercial tool. It is not part of the Digital Government Authority, is not endorsed by it, and does not act on its behalf. What it produces is an indicative estimate expressed as a range. It is not the Platforms Code Application Certificate and not a declaration of conformity. The certificate is issued by the Authority alone, and the Authority’s assessment remains the reference.

Its usefulness is in showing the gap early and with evidence. It does not stand in for the body that decides.