feat: add accessible Storybook UX contracts - #246
Conversation
|
Warning Review limit reachedNext included review available in 1 minute. View limit detailsLimit details: You’ve used the included review currently available. You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Review configuration: ⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (7)
📝 WalkthroughWalkthroughStorybook 접근성 설정과 Provider 상태 카드가 추가되었습니다. 디자인 토큰과 전역 접근성 스타일을 도입했습니다. 고객용 화면은 내부 오류와 구현 세부정보를 숨기도록 갱신했습니다. iCloud stall clock과 선호 스캔 루트를 추가했습니다. ChangesUX 및 플랫폼 변경
Estimated code review effort: 4 (Complex) | ~45 minutes Merge Risk: 🟡 Moderate · up to The PR currently has a compile-time contract error in iCloud stall handling and presents an incomplete description of the files Homebrew cleanup may remove. Merge should wait until the type/fallback behavior and cleanup-scope wording are corrected. Sequence Diagram(s)sequenceDiagram
participant GitHubActions
participant Storybook
participant Chromium
participant TestRunner
participant ProviderStatusCard
GitHubActions->>Storybook: 정적 빌드와 로컬 서버 시작
Chromium->>Storybook: 스토리 페이지 로드
TestRunner->>Chromium: 스토리 전역 viewport 적용
Chromium->>ProviderStatusCard: 상태 카드 렌더링
ProviderStatusCard-->>TestRunner: 접근성 및 취소 동작 결과 반환
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
Full details: Docstring CoverageExplanation Docstring coverage is 14.29% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 21 functions across 28 files. (12 skipped: 12 unsupported.) ✨ Finishing Touches 💡 2📝 Generate docstrings 💡
🛠️ Fix failing CI checks 💡
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
* feat: add ontology-bound orphan cache cleanup * docs: bind orphan baseline to live provider evidence * test: cover orphan cache and support roots * docs: record orphan rust test evidence * test: reject orphan object identity replacements * docs: bind orphan identity test evidence * docs: record repeated provider stall evidence * docs: bind baseline to latest orphan head * fix: fail closed on deep app inventory scans * fix: propagate incomplete orphan scan evidence * docs: bind orphan scan safety to current head * test: pin orphan cleanup authority and privacy regressions * fix: align orphan cleanup UI with fail-closed authority * fix: harden orphan planning authority and privacy * test: bind orphan cleanup regressions to repaired boundaries * test: prevent orphan plan fingerprints from encoding HOME * test: cover general reverse-DNS orphan bundle identities * fix: remove HOME-derived orphan authority metadata * test: require HOME-free orphan wire contract * fix: align orphan API with path-free runtime contract * test: require orphan batch preflight before mutation * fix: preflight orphan cleanup batches before mutation * fix: avoid recursive read directory binding * docs: refresh orphan current-head evidence * docs: distinguish orphan implementation head * test: separate orphan cleanup success from refresh failure * fix: preserve orphan cleanup success across refresh failure * fix: distinguish orphan cleanup from refresh failure * test: bind orphan refresh outcome to UI contract * docs: refresh orphan frontend test evidence * docs: record orphan refresh authority boundary * docs: record live iCloud transfer evidence * docs: track orphan safety follow-ups * docs: bind orphan audit to merged base head * docs: record orphan audit documentation head * docs: label orphan evidence snapshot * fix: harden orphan evidence and goal commands * fix: harden installed app inventory evidence * fix: fail closed on recursive lsof diagnostics * test: require orphan cleanup component-owned styles * fix: scope orphan cleanup presentation styles * test: decouple orphan bundle fixture from host Spotlight completeness * test: distinguish benign lsof warnings from target traversal failures * fix: classify benign lsof warning continuations * test: remove host-dependent Spotlight assertion * fix: bound orphan probes and revalidate metadata * docs: record orphan deadline and manifest revalidation * fix: terminate launch services descendants on timeout * docs: record launch services timeout isolation * docs: record exact-head stacked PR audit * docs: bind orphan audit to pushed head * docs: label orphan implementation snapshot
![devin-ai-integration[bot]](https://avatars.githubusercontent.com/in/811515?s=60&v=4){kind=small}
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
41fa888 to
eb89d67
Compare
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
| return [admissionFingerprint(report), progressFingerprint(report)].join("|"); | ||
| } | ||
|
|
||
| export function updateIcloudHealthStallClock( |
There was a problem hiding this comment.
🔍 Stall-clock module wired only to its test
updateIcloudHealthStallClock and its fingerprint helper are imported only by the test file; no component uses them. The logic ships as dead code. Confirm whether integration was intended here or deferred.
Was this helpful? React with 👍 or 👎 to provide feedback.
There was a problem hiding this comment.
Confirmed against the fresh branch: this module was no longer integrated by any production component after CloudArchive/provider ownership had already converged out of #246; it was imported only by its own test and did not exist on protected main. The root-cause repair is ownership convergence, not another provider workaround in the Storybook UX owner: icloudHealthStallClock.test.ts and icloudHealthStallClock.ts were removed from this branch in 3f4f878fc9dcbcbed8d2559e0ee8e26861937a5c and e1856c096c033d5afa492d9d367d6b09ca4fdbab. Exact-head checks are queued, so the thread stays unresolved pending terminal verification.
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
| function podmanActionLabel(action: { kind: string }): string { | ||
| return action.kind === "dangling-image" | ||
| ? "재생성 가능한 미사용 이미지 정리" | ||
| : "Podman 저장 공간 확인"; | ||
| } |
There was a problem hiding this comment.
🟡 Podman recommended actions all show one generic label
podmanActionLabel compares the action kind to "dangling-image", but the backend never emits that value; its kinds serialize as snake_case like review_unused_images. The first branch is dead, so every recommended action renders the same generic "Podman 저장 공간 확인" label instead of describing what it recommends.
Prompt for agents
In src/lib/Cleanup.svelte, podmanActionLabel checks action.kind === "dangling-image", but the recommended-action kind values coming from the Rust backend (PodmanRecommendedActionKind in src-tauri/src/podman_reclaim.rs, serialized snake_case) are review_unused_images, review_unused_volumes, restore_guest_headroom, investigate_api, review_guest_trim, and review_stopped_containers. Because none of these equals "dangling-image", the first branch is unreachable and every recommended action gets the generic fallback label. Map each real kind string to appropriate customer-facing copy (the "재생성 가능한 미사용 이미지 정리" label likely belongs to review_unused_images), so the displayed label matches the actual recommendation.
Was this helpful? React with 👍 or 👎 to provide feedback.
| }: Props = $props(); | ||
|
|
||
| const stateLabel: Record<ProviderStatusState, string> = { | ||
| clear: "지금 복사 가능", |
There was a problem hiding this comment.
🟡 Clear status overstates copy readiness
When provider observation is clear, stateLabel declares copying possible now despite the required per-file review. Users can proceed before the remaining safety decision.
Was this helpful? React with 👍 or 👎 to provide feedback.
seonghobae
added
the
enhancement
ChatGPT Codex Connector
2 participants
Purpose
Deliver the accessible Storybook/desktop UX contract while keeping provider, cleanup, release, scan-navigation, and destructive authority in their canonical owners.
Exact current state
main@79067c1160ddedf7fc962cbf8067ce7e83c4564a1a8ca96116f34720501a0f6c94e29dd48af2741dOwnership convergence in this head
A stale 58-file branch had accumulated release-verifier, release-version/provenance, canonical product-baseline, provider/cloud, TopFiles, and scan-page changes owned by other DiskSage lanes. The branch was converged non-destructively against current
mainand those foreign-to-this-PR deltas were restored to protected-main truth. In particular:TopFiles.svelteand the scan page returned to their dedicated owners;The remaining source is the Storybook/design-token/accessibility shell, preferred initial scan-root UX, focused customer-copy/error-feedback contracts, and the supporting provider-status presentation helpers.
Current product repair
A current review finding showed
Duplicates.svelterendering raw per-item backend errors ({r.error}) directly in customer markup. The repair was test-first:cleanupCustomerCopyContract.test.tsso Inventory/Duplicates may not render raw caught or per-item errors;The finding remains unresolved until the unchanged exact head obtains current hosted GREEN evidence.
UX / authority boundary
list_rootsTauri command keeps its external name while the UX preference lives inpreferred_scan_roots;/, avoiding accidental whole-filesystem-first scanning without removing explicit root access;Required before integration
Do not merge until this exact unchanged head has terminal-success applicable Test/Release/Security/SAST/accessibility and central required workflows, zero valid unresolved findings, the live review/ruleset requirements are satisfied, and repository-wide exact production coverage/package/SBOM/provenance requirements remain intact. Predecessor, pending, skipped-required, failed, stale, synthetic, status-only, or model-only evidence is non-passing.
Summary by CodeRabbit
새 기능
개선 사항
문서