Grabzies WCAG 2.2 Evidence Reconciliation

Prepared: 23 October 2026

Method: Reconciliation of existing saved reports and test records only. No new audit or browser tests were run.

Scope: https://www.grabzies.co.uk/ and the specific pages/behaviours identified in the cited records.

Executive conclusion

Full WCAG 2.2 Level A or AA conformance is not established by the evidence reconciled here. This is not a claim that every criterion fails, and it is not proof that no separate manual assessment exists elsewhere.

1. Keep the two report panels distinct

Report 1 — Automated Accessibility Test Results

The saved report snapshot records:

Measure	Recorded result
Automated score	50/100
Passed automated checks	19
Errors	2
Warnings	4
Score formula	`100 − (2 × 15) − (4 × 5) = 50`

This score is the report's own issue-weighted score. It is not a WCAG conformance percentage or certificate. The six individual findings include two detected failures and four items marked for investigation; they should not be treated as six failed WCAG success criteria.

Specific detected issues

WCAG 1.4.10 — Reflow: the saved rendered report records 336px page width at a 320px viewport and lists offending elements.
WCAG 1.4.11 — Non-text Contrast: eight controls were flagged for low-contrast edges, with sample ratios and colours. Applicability of a control boundary requires context-sensitive review.
Other recorded checks cover text contrast, keyboard focus, text spacing, 200% zoom, orientation, motion/flashing and selected multi-page consistency. These checks have scope limits; for example, 22 keyboard tab stops were sampled and two text elements could not be measured automatically.

2. Report 2 — Criterion-level WCAG 2.2 A/AA statuses

The saved criterion-level table reports all 55 Level A and AA criteria:

Status in saved report	Level A	Level AA	Total
Failed	0	2	2
Warning	1	0	1
No problems found	0	1	1
Manual review needed / partial automated evidence	30	0	30
Needs a person to check	0	20	20
Not checked this time	0	1	1
Total	31	24	55

The two reported failures are:

1.4.10 Reflow — Level AA.
1.4.11 Non-text Contrast — Level AA.

The one warning is 2.4.1 Bypass Blocks in the earlier criterion table. The table also reports 1.3.4 Orientation as “No problems found”; that label is limited to the implemented check and is not a complete conformance verdict.

The report's statement that there were zero automated review findings refers to a separate count of individual review findings. It does not cancel the two failures in Report 1 or the two failed criteria in Report 2, and it does not mean that all criteria passed.

3. Later evidence and what it changes

Six responsive browser checks

The existing remediation evidence records six checks that passed within their tested scope:

/shop/ at 320px, 375px and 480px.
/google-seo-audit/ at 320px, 375px and 480px.
Each recorded HTTP 200 and no horizontal overflow at the tested viewport width.

Interpretation: These results are valid evidence for those specific page/viewport combinations. They do not, by themselves, prove that the earlier 336px reflow finding is resolved on the exact same rendered page, content state and offending elements. They do not test the eight low-contrast control edges or all 55 criteria.

Later WCAG 2.4.1 regression record

The later evidence reconciliation records 5/5 passes for the WCAG 2.4.1 report-classification regression, superseding the earlier hidden-skip-link misclassification for that tested classification behaviour.

Interpretation: Record this as a later, scoped improvement to report classification. Do not rewrite the original report snapshot as though its warning never existed, and do not interpret the 5/5 classification regression as proof that every bypass mechanism works by keyboard and assistive technology or that the entire criterion conforms.

Other recorded engineering/regression results

The existing reconciliation also records:

Official ACT text-spacing examples: 38/38 matched, within the two scoped rules.
Text-spacing regressions: 30/30 passed, within the tested suite.
Corrected custom-audit fixtures: 4/4 detected, within those fixtures.
W3C readiness checks: 6/6 passed for publication/source integrity/recorded integration only.

These are tool/test-suite evidence. They are not site-wide WCAG conformance results for Grabzies, and ACT inventory counts or preliminary comparison counts must not be presented as verified pass totals.

4. Reconciliation rules applied

Preserve the 50/100 score and 19/2/4 counts as the saved Report 1 snapshot; do not reinterpret the score as conformance.
Preserve the 55-criterion table as the saved Report 2 snapshot, with its original status categories.
Add later evidence only to the exact behaviour and scope it retested.
Mark the earlier 2.4.1 warning as historically recorded and separately note the later 5/5 classification regression.
Treat the six responsive checks as scoped supplemental evidence, not proof that every earlier reflow issue is resolved.
Keep site-specific findings separate from AIWebPageSEO's own accessibility-tool/ACT regression tests.
Do not claim “55/55 passed”, full Level A/AA conformance, legal compliance or certification on the strength of these records.

Canonical wording for the report

The saved accessibility report covers all 55 WCAG 2.2 Level A and AA success criteria and provides criterion-level statuses. Its recorded results include two failed Level AA criteria (Reflow and Non-text Contrast), one warning, one criterion reported as having no problems found, 30 criteria requiring further review, 20 requiring a person to check, and one not checked this time. Subsequent evidence records six responsive checks passing within their tested scope and a 5/5 regression result for WCAG 2.4.1 report classification. These later results apply only to the behaviours and scopes tested; they do not establish full WCAG 2.2 Level A or AA conformance for Grabzies.

Source record trail

Saved report snapshot in the conversation/library: “Accessibility Audit — Two Evidence Reports”, including Report 1 and Report 2.
GRABZIES-WCAG-REMEDIATION-EVIDENCE.md
GRABZIES-WCAG-REMEDIATION-EVIDENCE.json
ACT-RECONCILED-EVIDENCE-2026-10-22.md
wcag-existing-evidence-reconciliation-20261023T160035Z.md
wcag-report-score-reconciliation-20261023T160244Z.txt
wcag-deep-dive-20261023T155238Z/deep-dive.md
wcag-report-findings-20261023T155355Z/findings.md

Limit: This reconciliation is based on saved conversation/library evidence, not a live inspection of the server. It does not modify the server's canonical files.