test(app): clean up v2026.5.11 release-gate e2e test debt by Astro-Han · Pull Request #569 · Astro-Han/pawwork

Astro-Han · 2026-05-12T06:01:02Z

Summary

Clean up the remaining #529 release-gate test debt without changing product behavior. Most of the original release-gate failures were already fixed on dev by earlier PRs; this PR removes the last stale session-composer-dock assumptions so future release checks fail only on real regressions.

Why

The original #529 bucket mixed current failures with assumptions that had already drifted out of date:

prompt-mention was already fixed on dev to search files in the active test workspace (5a787fc5e).
release-notes zh setup was already fixed on dev to pass LANGUAGE_KEY into browser init (8e6933dbf).
home.spec root-route behavior was already triaged as intended no-project empty state, not a regression (1e965b20d), and the model-chip assertion was already updated for intrinsic sizing (b25a0687f).
the stale session header-menu coverage had already been retired after the sidebar IA change (0e99d6c5f).

What was still live was inside session-composer-dock.spec.ts:

blocked question flow supports submitting after skipping every question still assumed the client needed an extra explicit Submit click after the final skip. Current behavior already auto-submits at that point, so the extra click was testing a stale contract.
e2e composer dock keeps latest turn visible when dock height changes depended on a nested-scroll / ongoing-rendering path that did not hold up as a stable user contract. It repeatedly behaved like release-gate debt rather than a trustworthy regression detector, so this PR retires it instead of teaching CI to chase a noisy path.

This PR is therefore test-debt cleanup, not a product fix.

Related Issue

Closes #529.

Human Review Status

Pending. A human should make the final merge decision after reviewing the final diff and verification evidence.

Review Focus

Whether removing the extra final Submit click correctly matches the current skip-all blocker contract.
Whether retiring the unstable dock-height test is the right boundary for #529, given that the path was noisy and not tied to a stable user-visible contract.
Whether the PR body draws the right line between already-landed fixes on dev and the remaining cleanup in this diff.

Risk Notes

No product behavior change. This PR only updates or retires stale E2E expectations.
Root route triage remains: / rendering the no-project empty state is intended IA drift, not a regression.
Header-menu rename/archive/delete coverage was already retired from the old path; current sidebar/session coverage now owns rename/delete behavior.
Windows advisory failures from the original release gate are out of scope for this PR and remain tracked separately.
During repeat verification I found an unrelated sidebar flake in sidebar-session-organization.spec.ts where the group-count assertion leaks state across repeat runs. It is not part of #529's changed surface, so it is tracked in a separate follow-up issue rather than expanded into this PR.

How To Verify

packages/app typecheck: pass
  bun --cwd packages/app typecheck

git diff check: pass
  git diff --check

Focused release-gate repeat run: 162 passed (4.8m)
  bun --cwd packages/app test:e2e \
    e2e/prompt/prompt-mention.spec.ts \
    e2e/release-notes/release-notes-toast.spec.ts \
    e2e/app/home.spec.ts \
    e2e/session/session-composer-dock.spec.ts \
    e2e/inputs/session-rename-dialog.spec.ts \
    --workers=1 --repeat-each=3

Screenshots or Recordings

Not needed. This is a test-only PR with no visible product change.

Checklist

Summary by CodeRabbit

Tests
- Improved test coverage for question dock interactions and visibility after submitting questions.

coderabbitai · 2026-05-12T06:01:14Z

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 696f322b-21a5-44cc-a203-46a53ba4ac1a

📥 Commits

Reviewing files that changed from the base of the PR and between 70488ee and c701d8b.

📒 Files selected for processing (1)

packages/app/e2e/session/session-composer-dock.spec.ts

💤 Files with no reviewable changes (1)

packages/app/e2e/session/session-composer-dock.spec.ts

📝 Walkthrough

Walkthrough

This PR removes test infrastructure debt from session-composer-dock.spec.ts by eliminating an unused filesystem import, deleting a scroll helper, refining skip-all-questions test flow, and replacing a stale scroll-height regression test with a dock-visibility assertion.

Changes

Composer Dock E2E Test Cleanup

Layer / File(s)	Summary
Unused imports and helpers cleanup `packages/app/e2e/session/session-composer-dock.spec.ts`	Removes unused `mkdir` import and deletes the `scrollTimelineAwayFromBottom` helper function.
Test flow and regression test updates `packages/app/e2e/session/session-composer-dock.spec.ts`	Updates skip-all-questions test to omit explicit submit-button click and assert dock opens directly; replaces "dock height changes" regression test with "submit to question dock keeps latest turn visible" test that validates dock and timeline bounding-box positions after question submission.

Possibly Related PRs

Astro-Han/pawwork#539: Directly related — both modify the same session-composer-dock E2E spec and handle scroll/dock test helpers and question-dock test adjustments.
Astro-Han/pawwork#387: Related — both adjust expectations around skip→submit and dock visibility behavior.
Astro-Han/pawwork#498: Related — both address session timeline auto-scroll behavior and dock visibility after submit.

Estimated Code Review Effort

🎯 2 (Simple) | ⏱️ ~8 minutes

Suggested Labels

flaky-test, P3, task

Poem

🐰 Hopping through the test debris so cluttered,
Sweeping away what's unused, no more fret—
Dock alignment shines through, assertion's uttered,
Cleaner E2E, no more scroll regret. ✨

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately describes the main change: cleaning up release-gate e2e test debt without product behavior changes.
Description check	✅ Passed	The description comprehensively covers all required sections: summary, why, related issue, human review status, review focus, risk notes, verification steps, and completed checklist.
Linked Issues check	✅ Passed	The PR addresses all in-scope objectives from `#529`: removed stale assumptions from session-composer-dock, documented already-landed fixes, and verified with targeted e2e repeat runs.
Out of Scope Changes check	✅ Passed	All changes are within scope as test-only cleanup in session-composer-dock.spec.ts. Unrelated sidebar flake was noted and tracked separately, not included.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

Create stacked PR
Commit on current branch

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch slock/release-gate-e2e-debt

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.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

Astro-Han

Opus second-pass review — PR #569

Pass, no blockers.

Strengths

Pure cleanup: 145 deletions, 0 additions, 1 file. No production code touched.
Both retired surfaces are well-justified in the PR body:
- The skip-all extra Submit click is genuinely stale; current contract auto-submits after the second skip and the regression test still asserts expectQuestionOpen(page) afterwards, so the user-visible behavior is still covered.
- The dock height changes test was the classic kind issue #529 flagged — "nested scroll / ongoing rendering" boundary rather than a stable user contract. Retiring it is the right call; chasing it further would have meant deeper harness coupling for what is effectively a flaky surface assertion.
Helper scrollTimelineAwayFromBottom and the mkdir import deletions are clean: both were only used by the retired test.
The "already fixed on dev" evidence chain for prompt-mention (5a787fc5e), release-notes zh (8e6933dbf), home.spec root-route (1e965b20d) is solid — this is exactly the kind of evidence-based scope discipline #529 asked for.
Stability proof is strong: --workers=1 --repeat-each=3 → 162 passed (4.8m) across all 5 spec files.
#570 follow-up cleanly carves out the unrelated sidebar-session-organization repeat-run group-count flake so it does not block this PR.

Verification confirmation

bun --cwd packages/app typecheck → pass
git diff --check → pass
Focused repeat run 162 passed across prompt-mention / release-notes-toast / home / session-composer-dock / session-rename-dialog

Verdict

Test-only cleanup, no product behavior change. Ready for GPT-X engineering final after first-pass and this review.

gemini-code-assist

Code Review

This pull request cleans up the E2E test suite by removing the test case 'e2e composer dock keeps latest turn visible when dock height changes' along with its associated helper function 'scrollTimelineAwayFromBottom' and an unused 'mkdir' import. Additionally, it removes a redundant submit button click from the blocked question flow test. I have no feedback to provide as there were no review comments.

test(app): clean up stale session composer dock cases

c701d8b

Astro-Han added P2 Medium priority app Application behavior and product flows tech-debt Supplemental cleanup, maintainability, architecture, test, or quality debt context labels May 12, 2026

Astro-Han mentioned this pull request May 12, 2026

[Bug] sidebar-session-organization.spec.ts group count assertion leaks state across repeat runs #570

Closed

Astro-Han commented May 12, 2026

View reviewed changes

Astro-Han merged commit 6689b6a into dev May 12, 2026
23 checks passed

gemini-code-assist Bot reviewed May 12, 2026

View reviewed changes

This was referenced May 12, 2026

feat(ci): add PR path-based label routing and priority recommendation #573

Merged

[Bug] test/file/index.test.ts flakes intermittently on Linux CI — empty result before walk completes #576

Closed

Astro-Han deleted the slock/release-gate-e2e-debt branch May 12, 2026 08:43

coderabbitai Bot mentioned this pull request May 21, 2026

fix(app): handle question response races clearly #839

Merged

13 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

test(app): clean up v2026.5.11 release-gate e2e test debt#569

test(app): clean up v2026.5.11 release-gate e2e test debt#569
Astro-Han merged 1 commit into
devfrom
slock/release-gate-e2e-debt

Astro-Han commented May 12, 2026 •

edited by coderabbitai Bot

Loading

Uh oh!

coderabbitai Bot commented May 12, 2026 •

edited

Loading

Walkthrough

Changes

Possibly Related PRs

Estimated Code Review Effort

Suggested Labels

Poem

Uh oh!

Astro-Han left a comment

Uh oh!

Uh oh!

gemini-code-assist Bot left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

Astro-Han commented May 12, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Why

Related Issue

Human Review Status

Review Focus

Risk Notes

How To Verify

Screenshots or Recordings

Checklist

Summary by CodeRabbit

Uh oh!

coderabbitai Bot commented May 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Possibly Related PRs

Estimated Code Review Effort

Suggested Labels

Poem

Uh oh!

Astro-Han left a comment

Choose a reason for hiding this comment

Opus second-pass review — PR #569

Strengths

Verification confirmation

Verdict

Uh oh!

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Astro-Han commented May 12, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented May 12, 2026 •

edited

Loading