Repository navigation
ci(ux-review): flag a painted scrollbar as a finding - #23568
Conversation
UX Review (Opus 5.5) — ⏭️ skippedRevision |
First Principles Review (Opus 5.5) — ✅ PASSPremise-level review of First-Principles-Verdict: PASS The rule is a recorded decision ( Notes
What this change shipsInventory (5 items) — 4 justifiedIntent: make the UX AI reviewer hold every frontend PR to the recorded "no scrollbar at rest" decision. ADDITION.
[FIRST-PRINCIPLES-REVIEWED] 915da2b Description read[DESCRIPTION-READ] 37a658893736c6036f02a40de5b75d5599705b2def30418a983483117105bccd That sha256 names the PR description this verdict read. Recompute it with |
Design Review (Opus 5.5) — ✅ PASSDesign-level review of Design-Verdict: PASS Nothing to check. This PR only adds prompt text, and every file, symbol and AUTOSDE rule it cites already exists on the base. [DESIGN-REVIEWED] 915da2b Description read[DESCRIPTION-READ] 48f0fd144dd717dd78c98edf16a6f93b44f957b72498b084cbc29ae6105314e9 That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with |
Opus 5.5 Review — ✅ no blocking findingsReviewed Verdict parsed from the review's SHA-scoped output markers for commit False positive or not applicable? A repository writer can comment: |
GPT 6.1 Review — ✅ no blocking findingsGPT 6.1 completed its review of This comment is updated in place on each push. Review detailsNo findings. False positive or not applicable? A repository writer can comment: |
859bb82 to
ed2f1d6
Compare
|
Bullet rewritten for the refined decision (maintainer, 2026-10-23): a scrollbar is never visible at rest; the two allowed forms are fully hidden and shown-only-while-scrolling without a width change. Hover-reveal is now named as a finding. The goal line was updated on that instruction; still draft until #23550 lands. |
ed2f1d6 to
3eed0fe
Compare
|
#23550 merged ( |
The dashboard shows no scrollbar at rest (maintainer decision, docs/decisions/2026-10-23-scrollbars-are-never-always-visible.md, landing in #23550): a box either hides it or shows a thin thumb only while it is scrolling, without changing its width. The UX lane had no lens for it: on #23550 it passed the change but could only note the capped lists that lost their hover thumb, with no rule to hold a future diff to. Lens 7 (consistency) now carries the rule in both UX lane prompts: a diff that paints a thumb at rest, a hover-revealed thumb, a reveal that pushes content, or a screenshot showing a scrollbar while nothing scrolls is a finding; a scroll box that needs a cue gets an edge fade or a "more below" affordance, never a resting thumb; hiding one is never a regression. Refs #23539
3eed0fe to
915da2b
Compare
iamwhatever
left a comment
There was a problem hiding this comment.
Approve. Reviewed 915da2b7d476e7612c76c201c4de2ebcdb0ef948.
The change adds the same lens-7 bullet to both UX lane prompts (ux-review.yml#L971-L993, fork-ux-review.yml#L975-L997). It is prompt text only: no permissions, secrets or steps change. Every file and identifier it cites (the decision record, lib/scrollReveal.ts, data-scrolling, .edge-fade-x, useScrollEdges, MORE_BELOW_MASK, JumpToBottomButton, the AUTOSDE rule) is on main now that #23550 has merged, which also settles the earlier ordering BLOCK. Current-head lanes are PASS or no findings, checks are green, and there are no change requests.
Optional: First Principles notes that the two CSS clauses duplicate scrollbarRevealOnly.test.ts; dropping them is a valid simplification but not required.
— crew-pr-reviewer (r127)
Problem / Motivation
Goal: The UX AI reviewer holds every frontend PR to the "no scrollbar is visible at rest" decision instead of passing one that paints a resting scrollbar back.
#23550 makes "no scrollbar is visible at rest; a box hides it or shows a thin thumb only while scrolling, without changing its width" one global rule, recorded in
docs/decisions/2026-10-23-scrollbars-are-never-always-visible.md. The UX Review lane had no lens for it: on #23550 it could only note that some capped lists lost their hover thumb, with no rule to judge the note by, and a later diff that reintroduced a thumb would meet no UX finding at all. The maintainer asked for the decision to become a UX reviewer rule (#23539).Why it matters
A design rule that lives only in a stylesheet and a CI test is enforced after the fact; the UX lane reads the diff and the screenshots and can name the cue a box should carry instead. Without the lens, every review of a new scroll surface re-litigates the decision.
Not a goal
website/src/test/scrollbarRevealOnly.test.ts) or theno-scrollbar-at-restAUTOSDE rule; both land in feat(dashboard): hide every scrollbar #23550.What changed (motivation → approach → change)
Decision → the lens that already holds house style (lens 7, "Consistency & habituation", where the flat-fills rule sits) → one new bullet in both UX lane prompts (
ux-review.ymlowns the text,fork-ux-review.ymlcarries the copy the parity test checks):scrollbar-none) or shows a thin thumb only while scrolling (data-scrolling), in a column it already reserves. A diff that paints one at rest is a finding: a coloured::-webkit-scrollbar-thumborscrollbar-colorwhose selector lacks[data-scrolling],scrollbar-width: auto, a hover-revealed thumb, a revived.scrollbar-overlay/.scroll-fade, a screenshot showing a scrollbar while nothing scrolls, or a reveal that pushes content..edge-fade-x,useScrollEdges) or a "more below" affordance (MORE_BELOW_MASK,JumpToBottomButton), never a resting thumb.The bullet cites the decision file by path, so once #23550 lands the lane can open it.
Backwards compatibility
Compatible: prompt text only. Removes nothing: no lens, marker or step is removed; the prompt stays expression-free and under the size the workflow header documents.
Tests
test/test_review_lane_prompt_parity.py(unchanged) pins that the lens block is identical across the two lanes; it covers the new bullet because the bullet sits inside the compared block.test/test_ai_review_workflows_ux.py(unchanged) covers the lane's step shape and markers.-n0).Manual verification
N/A — prompt-text change; the parity test and YAML parse are the verification, and the lane's behaviour on a real PR is what the next frontend PR shows.
Screenshots / video
N/A — no user-facing surface changes; this is a CI workflow prompt.
Related Issues
Refs #23539
Depends on #23550 (the decision file the lens cites)
Checklist
feat|fix|docs|style|refactor|perf|test|chore|ci|build|revert: ...)Contribution License Agreement
Contributions are accepted under the project's Apache-2.0 license; see the Licensing section of CONTRIBUTING.md.