Repository navigation
test(e2e): drive the undriven flows and run every journey in the merge queue - #151
Merged
Merged
Conversation
…hold notice of the request logs
Upd4ting
marked this pull request as ready for review
November 19, 2026 03:59
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
Wdf26acfc-17: every browser test of dms-api runs in the standard Playwright suite (
playwright test -c e2e, through the@antelopejs/dms/e2ekit), the@smokeones on every pull request and the full suite in the merge queue. Flows the verify feature map listed asNot drivennow have journeys, and the command-line checks of the verify harness run as ordinary Vitest tests on every pull request.Summary
Not drivenflows of.agents/skills/verify-dms-api/coverage/. Each one is listed in its feature file and infeatures/README.md, and its coverage row names it.SKILL.md's Not driven list shrinks to match.tests/package.test.ts, withtests/helpers/package.ts). They run the checks of the command-linepackagejourneys (npm pack, ranges, skill, docs links, locales) anddms-control map. Before this, those checks ran only nightly or in a verify run.semverand@types/semverbecome devDependencies for the range checks.ci.yml).selectoutputse2e:noneon a draft.smoke(--grep @smoke) on a pull request or a push.allonmerge_group,schedule,workflow_dispatchand a pull request labelledci:full. It runs two shards plus the merged report, and theubuntugate requires the job whenevere2eis notnone.merge_grouptrigger belongs to theon:block thatantelopejs-syncowns. It comes with the move to@antelopejs/tooling-configs>=0.8.9 <0.9.0(root andfrontend-vue). That sync also moves the push trigger and the release workflow tomain, removesnext-sync.ymlandstale.yml, and turns onaffected-tests.@smoke. The 7 existing tags stay: one per feature, withsmokefirst in the feature map. The new journeys carry none.Feature-map flows and the spec that covers each
New journeys, all in the Playwright suite:
Not driven)/opens the search,?lists the shortcutsslash-opens-the-request-search-and-question-mark-lists-the-shortcutse2e/journeys/logs/keyboard-and-drawer.mjscopy-curl-copies-the-open-request-as-a-commanddragging-the-drawers-left-edge-widens-it-and-it-opens-again-at-that-widthan-open-row-or-an-active-search-holds-the-table-and-says-why(oneknownFailuresentry, below)Ltoggles Live and Pausel-switches-the-page-between-live-and-pausee2e/journeys/overview/live-shortcut.mjs/opens the route filterslash-opens-the-route-filtere2e/journeys/routes/keyboard-and-clipboard.mjscopy-curl-on-a-route-copies-an-example-callthe-use-default-button-removes-an-override-on-the-pagee2e/journeys/settings.mjsa-member-without-the-modules-permissions-is-sent-to-their-settingse2e/journeys/access.mjsa-panel-opened-from-one-table-reads-only-a-request-that-table-can-listthe-forwarding-route-refuses-unsafe-paths-and-never-forwards-blocked-headerse2e/journeys/tester/forwarding.mjsThe other flows of the feature map were already covered and keep their journeys unchanged:
smoke,capture,overview,logs,routes,tester,settings,retention,accessandpackage(e2e/journeys/<feature>.mjs, listed per feature in.agents/skills/verify-dms-api/features/README.md).Still not driven, with the reasons in
coverage/: the automation runtime's election, a second backend sharingtesterSourceSecret, the Tester's own Copy cURL and editors, the 30 s forwarding timeout, and the Copy buttons beside a value.Non-browser checks and the tests that run them
dms-controltests/package.test.ts)packagejourneythe-tarball-holds-what-the-manifest-names(npm pack --dry-run)the packed tarball › holds the built entry…,› holds nothing outside dist…packagejourneythe-ranges-and-the-skill-match-what-the-docs-namethe ranges the docs name › …(3),the skill › …(2)packagejourneythe-docs-links-and-the-two-locales-are-completethe docs and the locales › …(2)dms-control mapthe verify feature map › covers every user-facing file…dms-control formal(TLC)formal-route-statisticsjob already runs it on every pull requestThe HTTP-only journeys (
access,capture,overviewverdicts,automation, and the new tester and access ones) stay in the Playwright suite. They need the DMS backend, MongoDB and auth that the kit's run starts, and the Vitest suite has no backend. Thepackagejourneys also stay, as the brief keeps every journey.Proof
Playwright test counts,
TZ=UTC pnpm exec playwright test -c e2e --list | tail -1:--grep @smokemain, 5b2dc87)Total: 74 tests in 1 fileTotal: 7 tests in 1 fileTotal: 85 tests in 1 fileTotal: 7 tests in 1 fileDMS_E2E_PREBUILT=1 TZ=UTC pnpm e2e -g "<ids>", and the three keyboard journeys passed twice more after a race fix.pnpm exec vitest run tests/package.test.tsgivesTests 10 passed (10).src/zz/probe.tsmakesdms-control mapprintunmapped: src/zz/probe.ts matches no feature globand exit 1. A feature naming an undescribed journey makes it exit 1 too.e2e (1)ande2e (2)passed, about 7 min each. Its one failure,sync:checkon themerge_grouptrigger, is fixed in 68ff39c.e2e (1)passed in 6.5 min ande2e (2)in 8.5 min, and so dide2e-report. Its one failure,size(e2e/journeys/tester.mjs1,364 bytes over 20,000), is fixed by moving the forwarding journey toe2e/journeys/tester/forwarding.mjs, which passed again locally.flowchart LR E[event] --> S{select} S -->|draft| N[e2e skipped] S -->|pull_request, push| SM["pnpm e2e --grep @smoke"] S -->|merge_group, schedule, workflow_dispatch, ci:full label| A["pnpm e2e --shard=i/2 + merged report"]Known failure
an-open-row-or-an-active-search-holds-the-table-and-says-whylistsan open request says the table is on hold, row openinknownFailures. The request panel is a modaldrawertarget, and the DMS names therow openhold only for a dockedsidePaneltarget. An open request does hold the table, behind the1 new request · Showpill, but never says why.docs/4.usage/1.overview.md(Live mode) promises the notice.Follow-ups
drawertarget holds its rows silently while a row is open, done means it showsTable on hold · row openas asidePaneltarget does, and the dms-api journey'sknownFailuresentry readsfixed