Carlos Florêncio
5.6 KiB
id, title, status, wave, depends_on, plan, requirements, acceptance_criteria, system_design
| id | title | status | wave | depends_on | plan | requirements | acceptance_criteria | system_design | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 03-removal-e2e | Prove removal transitions | done | 3 |
|
plan.md |
|
|
|
Task 03: Prove removal transitions
Summary
Prove the entire removal interval on desktop and phone using controlled transport ordering. Retain the existing destination, cascade, and confirmation coverage while adding assertions that detect transient outgoing UI.
In scope
- Execute the plan's E2E matrix with isolated managed backend fixtures.
- Add a mobile delete scenario file and shared observation/response-gate helpers only where multiple scenarios need them.
- Observe outgoing mounts and forbidden state changes from acceptance through departure; assert before releasing delayed responses.
- Verify no outgoing ensure request, no main-frame reload, matching URL/content, later navigation, preview closure, bulk partial failure, and failure restoration.
- Inspect a rendered phone status/destination and record focus and layout evidence.
- Record new results and synchronize this package's statuses after all checks pass.
Out of scope
Broad E2E/full verification, generic QA/review, backend behavior changes, rewriting historical companion-package test counts, commit/push/PR operations.
Acceptance
- Both projects prove the outgoing task cannot display teardown/recovery UI during a pending request, including the last-task path.
- Failure, cascade, batch, preview, and later-navigation scenarios exercise user controls and assert intermediate as well as final state.
- All listed commands pass with discovered test counts and exact results recorded; every affected AC has executable evidence and the phone rendering is inspected.
Verification
Run from the repository root, sequentially. Managed commands rebuild the product. Do not overlap them or use additional worker overrides.
(cd apps/web && rtk pnpm e2e:run --project chromium tests/task/delete-task-redirect.spec.ts tests/task/archive-task-redirect.spec.ts tests/kanban/card-menu-delete-archive.spec.ts tests/task/sidebar-multi-select.spec.ts tests/task/archive-confirmation-preference.spec.ts)
(cd apps/web && rtk pnpm e2e:run --project mobile-chrome tests/task/mobile-delete-task-redirect.spec.ts tests/task/mobile-archive-task-redirect.spec.ts tests/task/mobile-archive-confirmation-preference.spec.ts tests/kanban/mobile-card-archive-confirmation.spec.ts)
rtk python3 scripts/lint-spec-files.test.py
rtk python3 scripts/lint-spec-files.py --all
rtk git diff --check
rtk git status --short -- docs/plans/task-removal-navigation
Confirm the runner discovers every intended file; zero tests is not success. If an implementation correction is needed, rerun its 01/02 checks and affected E2E files after the final correction. Use the recorded component RED from 02 when the browser fixture cannot independently establish pre-fix ordering; explain that limitation rather than claim a browser RED that was not run.
Files likely touched
- All E2E files in the plan's scenario table
apps/web/e2e/tests/task/mobile-delete-task-redirect.spec.ts(new)apps/web/e2e/tests/task/removal-transition-helpers.ts(proposed shared helper)apps/web/e2e/pages/session-page.tsand phone page helpers as needed- This plan/work orders and paired spec lifecycle fields
Dependencies
02 supplies rendered integration and its unit/component RED/GREEN evidence.
Risks
Polling only for a final URL can pass despite the original flicker. Install
observation before the action and explicitly assert during controlled pending
HTTP and after correlated WS events. Release gates and dispose observers in
finally, including failed assertions. Avoid helper methods that silently reload
or resume a session to make the test pass.
Parallelism
sequential
Inputs
- Full plan E2E matrix and paired requirements.
- Existing archive/delete redirect and mobile archive specs.
- E2E skill, causal-wait helpers, session page objects, and mobile-parity guide.
- 01/02 recorded outcomes.
Results
Done on 2026-09-10.
- Chromium removal matrix: 23 tests passed.
- Mobile Chromium removal matrix: 5 tests passed.
- Coverage includes active/last-task archive and delete, cascade exclusion, held mutation responses, outgoing DOM protection, preview actions, bulk selection, and desktop/mobile archive confirmation behavior.
- The delete page-object waits for the coordinator's localized success toast before starting a second removal; the held-response scenario opts out so it can inspect the pending interval.
- Inspected the phone rendering in
apps/web/.pr-assets/mobile-pr-removal-capture--mobile-task-removal-status.png; the captured surface shows the neutral removal status in the existing mobile content region. Mobile E2E asserted the final task/overview destination, URL identity, and absence of a document reload after the held response was released. - Specification tests: 36 passed; full specification lint passed.
- Targeted E2E-sleep lint for changed files and
rtk git diff --checkpassed. - The repository-wide E2E-sleep config still reports unrelated baseline errors; the changed E2E files are clean under the same config.