[lidge-jun]

Asking an agent "does this page meet WCAG 2.2 AA?" got contrast ratios and a 44 px rule from scattered files, with no owner for the 55 A/AA criteria, nothing on the six criteria new in 2.2, and Lighthouse scores treated as a pass. Fixing a visual audit had no per-violation verifier or before/after contract either. This PR adds both owners to dev-frontend.

Stack: 2 of 3. Base docs/frontend-skill-audit-uiux (#327). Next: #329.

Before / After: the documented check on a seeded page

The same pagination fixture, 390 px wide. Before: two 12 × 12 px links and an image without alt. After: 44 px targets with spacing and an alt text.

Before	After
[{"id":"image-alt","nodes":1},{"id":"target-size","nodes":2}] exit 1	[] exit 0

Output of the script in wcag22-aa.md (axe-core 4.14.0, tags wcag2a,wcag2aa,wcag21a,wcag21aa,wcag22aa), saved in evidence/axe-fixture/run.log. target-size is WCAG 2.5.8, one of the 2.2 additions.

What changed

Need	Before	After
Conformance answer	No owner; "WCAG AA minimum" meant contrast	wcag22-aa.md: 55 criteria (31 A, 24 AA) with axe rules or a manual check each; FE-WCAG22-01 (STRICT) forbids conformance claims from automated runs, scores or screenshots
2.2 additions	Only target size and focus mentioned	2.4.11, 2.5.7, 2.5.8, 3.2.6, 3.3.7, 3.3.8 with pass evidence; 4.1.1 marked obsolete
Running axe	Not specified	Playwright script with a scratch install; two observed traps written down (CLI exits 2 on a ChromeDriver/Chrome mismatch; AxeBuilder needs browser.newContext())
Fixing violations	Screenshot list per viewport	visual-audit-loop.md: violation ledger (V-ids carrying GAP/EV ids), verifier and pass rule written before the fix (FE-VISUAL-AUDIT-02), recapture after the last edit, closure gate (03), PR Before/After section, cxc-loop routing with --surface web criteria

Routing rows in SKILL.md and topic-index.md; pointers in role-boundaries, visual-verification, preflight-full and anti-slop-enforcement.

Validation

• Gate, prompt-architecture and the five skill tests exit 0 (46/46) through cxc receipt test.
• The fixture run above, repeated by two independent reviewers from their own copies.
• A fresh reviewer ran the documented script against the fixture and traced a GAP/EV id to a V row and the PR section: PASS for both criteria.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-27T16:53:17.636284Z	50b1181	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 50b118143a

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Separate KWCAG conformance from the WCAG 2.2 checklist

When the user asks for KWCAG conformance, this instruction routes the agent to a WCAG 2.2 AA checklist and its verdict rules as though the standards were interchangeable. KWCAG 2.2 is a separate Korean standard based on WCAG 2.1, with its own 33 inspection items, rather than the 55 criteria documented here. See the official standard listing and the government's description of its 33-item assessment. Remove the implied equivalence and require a KWCAG-specific checklist or explicit mapping before issuing a KWCAG verdict; otherwise an agent can complete this checklist without establishing the conformance the user actually requested.

Useful? React with 👍 / 👎.

[lidge-jun]

Agreed, fixed in ed63835. The opening line no longer calls KWCAG something that follows WCAG 2.2, and FE-WCAG22-03 (STRICT) now says this checklist does not decide KWCAG: KWCAG 2.2 (KS X OT0003, 2022) is built on WCAG 2.1 with 33 inspection items, so a KWCAG verdict needs that item list or an explicit item-by-item mapping that names unmapped items. Merged up into #329 (9e68ba5).

[github-actions]

✅ Target branch corrected

This pull request now targets dev.

The [WRONG BRANCH] title prefix has been removed. Its existing draft status has been preserved.