Repository navigation
test: reorganize subgraph test suite into composable domain specs - #10759
Conversation
Create the new domain-oriented Playwright suite under browser_tests/tests/subgraph so the remaining refactor can proceed from stable file ownership instead of the legacy flat subgraph specs. What this commit includes: - add the first-pass browser suite split into dedicated domain files for slots, promotion, promotion DOM, serialization, navigation, nested behavior, lifecycle, CRUD, search, and in-subgraph operations - preserve the migrated browser coverage in the new folder without deleting the legacy flat specs yet, which keeps later parity checks and matrix-thinning work straightforward - keep the helper surface narrow by moving organization into spec files instead of hiding broad assertions behind new helper APIs - strengthen SubgraphNode unit coverage so slot rename semantics and label propagation have an explicit lower-level owner before more Playwright thinning happens - add store-level coverage for subgraph navigation workflow restoration and unreachable-subgraph fallback so browser navigation tests can stay focused on visible flows in later batches Why this checkpoint matters: - Batch 1 needed the browser suite reorganized by domain before semantic pruning could happen safely - Batch 2 needed lower-level ownership established for rename/label propagation and workflow navigation policy before browser coverage was reduced further - this commit intentionally stops before Batch 3 because promotion and serialization still share ownership-heavy files and will be handled in a single-owner follow-up pass Notes on current ownership after this commit: - browser_tests/tests/subgraph/subgraphSlots.spec.ts now acts as the browser home for slot and rename journeys - src/lib/litegraph/src/subgraph/SubgraphNode.test.ts owns the renamed label propagation cases that should not stay duplicated in Playwright - browser_tests/tests/subgraph/subgraphNavigation.spec.ts carries the UI navigation journeys while src/stores/subgraphNavigationStore.test.ts owns workflow-path restoration and orphan-subgraph fallback policy - promotion and serialization specs are present in their new domain files but have not yet been matrix-thinned Verification run for this snapshot: - pnpm exec vitest run src/lib/litegraph/src/subgraph/SubgraphNode.test.ts src/stores/subgraphNavigationStore.test.ts - pnpm typecheck:browser - pnpm exec eslint browser_tests/tests/subgraph src/lib/litegraph/src/subgraph/SubgraphNode.test.ts src/stores/subgraphNavigationStore.test.ts - pnpm exec oxlint browser_tests/tests/subgraph - pnpm typecheck - pnpm exec oxfmt --check browser_tests/tests/subgraph src/lib/litegraph/src/subgraph/SubgraphNode.test.ts src/stores/subgraphNavigationStore.test.ts Amp-Thread-ID: https://ampcode.com/threads/T-019d3fa7-8653-71c8-8555-dae0c38324a2 Co-authored-by: Amp <amp@ampcode.com>
Amp-Thread-ID: https://ampcode.com/threads/T-019d4031-7021-74d9-b3c3-118f771fb064 Co-authored-by: Amp <amp@ampcode.com>
Amp-Thread-ID: https://ampcode.com/threads/T-019d4065-bee5-7118-8a5d-64cd3b998df5 Co-authored-by: Amp <amp@ampcode.com>
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
📝 WalkthroughWalkthroughConsolidates and tightens many subgraph Playwright and unit tests: adds serialization and duplicate-ID tests, rehomes helpers, reduces/merges redundant scenarios, strengthens async assertions with polling, and updates selectors and Comfy feature flags across test suites. Changes
Estimated code review effort🎯 4 (Complex) | ⏱️ ~50 minutes Possibly related PRs
Suggested labels
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
There was a problem hiding this comment.
Actionable comments posted: 10
🧹 Nitpick comments (4)
browser_tests/tests/subgraph/subgraphNested.spec.ts (1)
93-135: Extract duplicated promoted-value checks into a helperLines 93-109 and Lines 119-135 repeat the same selector/assertion flow. A small helper will reduce drift and keep this test easier to maintain.
As per coding guidelines, "Use Refactoring to make complex code simpler" and "Whenever new code is written, ask if there's a simpler way to introduce the same functionality; choose the simpler course."
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphNested.spec.ts` around lines 93 - 135, Extract the duplicated assertion block into a small helper (e.g., assertPromotedValues or checkPromotedNodeValues) that accepts a node locator (the objects returned by comfyPage.vueNodes.getNodeLocator or nodeLocator.getByLabel/.getByRole) and performs the four checks: getInputNumberControls for 'width', 'height', and 'steps' and getByRole('textbox', { name: 'prompt' }) then assert the inputs have values '1024', '1024', '8' and the prompt matches /Latina female/; replace both repeated blocks (the one using the original nodeLocator and the one using nodeAfter) with calls to this helper, and keep using comfyPage.vueNodes.getInputNumberControls and nodeLocator/getNodeLocator inside the helper to locate controls.browser_tests/tests/subgraph/subgraphCrud.spec.ts (1)
20-27: Guard Alt key release withtry/finallyin drag helper.If an action throws before Line 26, Alt can remain pressed and leak state into later tests.
♻️ Suggested hardening
await comfyPage.page.mouse.move(subgraphPos.x + 16, subgraphPos.y + 16) await comfyPage.page.keyboard.down('Alt') - await comfyPage.page.mouse.down() - await comfyPage.nextFrame() - - await comfyPage.page.mouse.move(subgraphPos.x + 64, subgraphPos.y + 64) - await comfyPage.page.mouse.up() - await comfyPage.page.keyboard.up('Alt') + try { + await comfyPage.page.mouse.down() + await comfyPage.nextFrame() + await comfyPage.page.mouse.move(subgraphPos.x + 64, subgraphPos.y + 64) + await comfyPage.page.mouse.up() + } finally { + await comfyPage.page.keyboard.up('Alt') + }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphCrud.spec.ts` around lines 20 - 27, The drag helper currently presses Alt via comfyPage.page.keyboard.down('Alt') but may not release it if an exception occurs; wrap the sequence that includes comfyPage.page.keyboard.down('Alt'), comfyPage.page.mouse.down(), the mouse.move/up calls, and comfyPage.nextFrame() in a try/finally so comfyPage.page.keyboard.up('Alt') is always executed in the finally block (ensure comfyPage.page.mouse.up() is also called or moved into finally if needed) — locate the calls to comfyPage.page.keyboard.down/up, comfyPage.page.mouse.down/up, and comfyPage.page.mouse.move in the drag helper and restructure them into try { ... } finally { await comfyPage.page.keyboard.up('Alt') } to prevent Alt state leaking between tests.browser_tests/tests/subgraph/subgraphSlots.spec.ts (1)
410-411: Consider documenting the double nextFrame intent.Two consecutive
nextFrame()calls after workflow load is presumably waiting for multiple render cycles. A brief comment would help future maintainers understand this is intentional, not accidental duplication.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts` around lines 410 - 411, Add a short inline comment explaining why two consecutive calls to comfyPage.nextFrame() are intentional (e.g., to wait for two separate render/update cycles after workflow load) so future maintainers understand this is deliberate; locate the calls to comfyPage.nextFrame() in subgraphSlots.spec.ts and add a one-line comment above them stating the specific reason (e.g., "wait for initial frame render and subsequent layout pass") and keep the calls as-is.src/lib/litegraph/src/subgraph/SubgraphSerialization.test.ts (1)
526-527: Replaceas string[][]cast with type guard for stricter fixture validation.The
Array.isArray()check only narrows tounknown[], leaving the subsequentas string[][]cast unchecked. Since the code only accessesentry[0]and converts it to string, use a type guard to validate the structure and improve type safety without relying on broad assertions.Suggested change
+function isProxyWidgetEntry(value: unknown): value is [string | number, ...unknown[]] { + return ( + Array.isArray(value) && + value.length > 0 && + (typeof value[0] === 'string' || typeof value[0] === 'number') + ) +} @@ - for (const entry of rootProxyWidgetsA as string[][]) { + expect(rootProxyWidgetsA?.every(isProxyWidgetEntry)).toBe(true) + for (const entry of (rootProxyWidgetsA ?? []).filter(isProxyWidgetEntry)) { expect(subgraphAIds.has(String(entry[0]))).toBe(true) } @@ - for (const entry of rootProxyWidgetsB as string[][]) { + expect(rootProxyWidgetsB?.every(isProxyWidgetEntry)).toBe(true) + for (const entry of (rootProxyWidgetsB ?? []).filter(isProxyWidgetEntry)) { expect(subgraphBIds.has(String(entry[0]))).toBe(true) }Applies to lines 526–527 and 532–533.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/lib/litegraph/src/subgraph/SubgraphSerialization.test.ts` around lines 526 - 527, The loop over rootProxyWidgetsA uses an unsafe cast "as string[][]"; replace that with a runtime type guard: iterate over rootProxyWidgetsA, check Array.isArray(entry) and that entry.length > 0 and typeof entry[0] === 'string' before using entry[0], then assert subgraphAIds.has(entry[0]) (or String(entry[0]) if you still need coercion); apply the same guard to the other occurrence (the loop at the later lines) to remove the broad cast and ensure strict fixture validation for rootProxyWidgetsA and its entries.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@browser_tests/tests/subgraph/subgraphCrud.spec.ts`:
- Around line 65-73: Replace the brittle loadWorkflow('default') +
getNodeRefById('5') pattern with a deterministic fixture: call
comfyPage.workflow.loadWorkflow(...) with the dedicated subgraph CRUD JSON
fixture from assets (e.g., the premade subgraph CRUD workflow) and locate the
node via a stable identifier (title or test id) using the nodeOps lookup helper
instead of getNodeRefById('5'); then call node.convertToSubgraph() as before.
Specifically update the test "Can create subgraph from selected nodes" to use
comfyPage.workflow.loadWorkflow('<assets fixture name>') and a lookup like
comfyPage.nodeOps.getNodeByTitle(...) or getNodeByTestId(...) in place of
getNodeRefById so the test no longer depends on the mutable default graph and
numeric IDs.
In `@browser_tests/tests/subgraph/subgraphLifecycle.spec.ts`:
- Around line 51-64: Add a precondition that the preview DOM is actually
rendered before deleting node '5': after obtaining beforePseudo via
getPseudoPreviewWidgets(comfyPage, '5') and before calling
subgraphNode.delete(), assert the page locator identified by
DOM_PREVIEW_SELECTOR (or comfyPage.page.locator(DOM_PREVIEW_SELECTOR)) is
present/visible (e.g., expect(...).toBeVisible() or
expect(...).toHaveCount(beforePseudo.length > 0 ? beforePseudo.length : 1)) so
the test guarantees the DOM-cleanup path is exercised; keep the existing checks
(getPseudoPreviewWidgets, subgraphNode.exists, countGraphPseudoPreviewEntries)
and insert this DOM assertion immediately before await subgraphNode.delete().
In `@browser_tests/tests/subgraph/subgraphNavigation.spec.ts`:
- Around line 8-33: The helper hasVisibleNodeInViewport currently returns true
for any partial overlap; change its logic (in the function
hasVisibleNodeInViewport iterating canvas.graph._nodes and using
node.pos/node.size and visLeft/visTop/visRight/visBottom) to require a full fit
by checking that nx >= visLeft AND nx + nw <= visRight AND ny >= visTop AND ny +
nh <= visBottom for at least one node; alternatively, if the test should assert
that all nodes fit, replace the per-node check with a combined-bounds check
computed from all nodes' positions/sizes against the visible rect.
- Around line 217-230: The test currently bypasses the real subgraph-enter flow
by calling canvas.setGraph() inside comfyPage.page.evaluate(), so replace that
direct mutation with a user-driven transition: inside the page.evaluate() find
the subgraph node (sgNode) as before but call the viewport entry helper (prefer
navigateIntoSubgraph() or canvas.openSubgraph()) instead of canvas.setGraph(),
ensuring the test drives the same entry path exercised by user interactions;
update any references to sgNode/subgraph accordingly so the test awaits the
navigation completion rather than forcing the graph state.
In `@browser_tests/tests/subgraph/subgraphNested.spec.ts`:
- Line 31: The current assertion expect(response.status()).not.toBe(400) is too
loose; replace it with an explicit success check on the Playwright Response
object—either expect(response.status()).toBe(200) if the endpoint should return
200, or use expect(response.ok()).toBeTruthy() to assert a successful response;
update the assertion that references response.status()/response.ok() in the test
in subgraphNested.spec.ts accordingly.
In `@browser_tests/tests/subgraph/subgraphOperations.spec.ts`:
- Around line 35-39: The test uses platform-specific keyboard shortcuts
'Control+c' and 'Control+v' which will fail on macOS; update the calls to
comfyPage.page.keyboard.press to use Playwright's cross-platform modifier
'ControlOrMeta+c' and 'ControlOrMeta+v' (i.e., change the two
comfyPage.page.keyboard.press invocations) so copy/paste works on macOS and
other platforms, keeping the following sequence:
comfyPage.page.keyboard.press('ControlOrMeta+c'), comfyPage.nextFrame(),
comfyPage.page.keyboard.press('ControlOrMeta+v'), comfyPage.nextFrame().
In `@browser_tests/tests/subgraph/subgraphPromotion.spec.ts`:
- Around line 175-225: Before each widget context-menu interaction select the
node title first: when you have ksampler (from getNodeRefById('1')) call
ksampler.click('title') before right-clicking the steps widget (stepsWidget) so
the menu target resolves; likewise, after navigating into the subgraph
(subgraphNode2.navigateIntoSubgraph()) call the inner nodeRef.click('title') for
the node you retrieved (the nodeRef used to get stepsWidget2) before
right-clicking widgetPos2 to trigger the Un-Promote Widget menu.
In `@browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts`:
- Around line 112-123: The teardown uses immediate snapshot counting via
comfyPage.page.locator(DOM_WIDGET_SELECTOR).count(), which can be flaky after
await subgraphNode.delete(); replace this with the retrying Playwright assertion
pattern used elsewhere: use comfyPage.page.locator(DOM_WIDGET_SELECTOR) combined
with expect(...).toHaveCount(0) so the check waits/retries until the widget is
removed; keep the same DOM_WIDGET_SELECTOR and call sequence
(getNodeRefById('11') then subgraphNode.delete()) but swap the
finalCount/count-based expectation for toHaveCount(0).
- Around line 126-149: The test currently only asserts internal widget count via
getPromotedWidgetCount('11'); additionally assert the actual DOM element is gone
by querying the page for the promoted widget selector (e.g. use
comfyPage.page.locator('.comfy-multiline-input' or DOM_WIDGET_SELECTOR) and
assert its count is 0) after calling comfyPage.subgraph.removeSlot('input',
'text') and comfyPage.subgraph.exitViaBreadcrumb(); place this DOM cleanup
assertion alongside the existing expect.poll(...) so the test verifies both
internal state (getPromotedWidgetCount) and that the actual
.comfy-multiline-input element has been removed from the DOM.
In `@src/lib/litegraph/src/subgraph/SubgraphSerialization.test.ts`:
- Around line 485-487: The test registers a global node type via
LiteGraph.registerNodeType('dummy', DummyNode) which pollutes the shared
registry; modify the test to isolate that registration by moving the register
call into a test-scoped setup (e.g. inside the it or a beforeEach) and ensure
you remove it after the test using LiteGraph.unregisterNodeType('dummy') (or a
finally/afterEach cleanup) so the global registry is restored; refer to
LiteGraph.registerNodeType, LiteGraph.unregisterNodeType and DummyNode when
making the change.
---
Nitpick comments:
In `@browser_tests/tests/subgraph/subgraphCrud.spec.ts`:
- Around line 20-27: The drag helper currently presses Alt via
comfyPage.page.keyboard.down('Alt') but may not release it if an exception
occurs; wrap the sequence that includes comfyPage.page.keyboard.down('Alt'),
comfyPage.page.mouse.down(), the mouse.move/up calls, and comfyPage.nextFrame()
in a try/finally so comfyPage.page.keyboard.up('Alt') is always executed in the
finally block (ensure comfyPage.page.mouse.up() is also called or moved into
finally if needed) — locate the calls to comfyPage.page.keyboard.down/up,
comfyPage.page.mouse.down/up, and comfyPage.page.mouse.move in the drag helper
and restructure them into try { ... } finally { await
comfyPage.page.keyboard.up('Alt') } to prevent Alt state leaking between tests.
In `@browser_tests/tests/subgraph/subgraphNested.spec.ts`:
- Around line 93-135: Extract the duplicated assertion block into a small helper
(e.g., assertPromotedValues or checkPromotedNodeValues) that accepts a node
locator (the objects returned by comfyPage.vueNodes.getNodeLocator or
nodeLocator.getByLabel/.getByRole) and performs the four checks:
getInputNumberControls for 'width', 'height', and 'steps' and
getByRole('textbox', { name: 'prompt' }) then assert the inputs have values
'1024', '1024', '8' and the prompt matches /Latina female/; replace both
repeated blocks (the one using the original nodeLocator and the one using
nodeAfter) with calls to this helper, and keep using
comfyPage.vueNodes.getInputNumberControls and nodeLocator/getNodeLocator inside
the helper to locate controls.
In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts`:
- Around line 410-411: Add a short inline comment explaining why two consecutive
calls to comfyPage.nextFrame() are intentional (e.g., to wait for two separate
render/update cycles after workflow load) so future maintainers understand this
is deliberate; locate the calls to comfyPage.nextFrame() in
subgraphSlots.spec.ts and add a one-line comment above them stating the specific
reason (e.g., "wait for initial frame render and subsequent layout pass") and
keep the calls as-is.
In `@src/lib/litegraph/src/subgraph/SubgraphSerialization.test.ts`:
- Around line 526-527: The loop over rootProxyWidgetsA uses an unsafe cast "as
string[][]"; replace that with a runtime type guard: iterate over
rootProxyWidgetsA, check Array.isArray(entry) and that entry.length > 0 and
typeof entry[0] === 'string' before using entry[0], then assert
subgraphAIds.has(entry[0]) (or String(entry[0]) if you still need coercion);
apply the same guard to the other occurrence (the loop at the later lines) to
remove the broad cast and ensure strict fixture validation for rootProxyWidgetsA
and its entries.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: e8570dac-4a95-4f2f-8cab-4057b0049c96
📒 Files selected for processing (30)
browser_tests/tests/subgraph-duplicate-ids.spec.tsbrowser_tests/tests/subgraph-promoted-slot-position.spec.tsbrowser_tests/tests/subgraph-promoted-widget-dom.spec.tsbrowser_tests/tests/subgraph-rename-dialog.spec.tsbrowser_tests/tests/subgraph.spec.tsbrowser_tests/tests/subgraph/subgraphCrud.spec.tsbrowser_tests/tests/subgraph/subgraphLifecycle.spec.tsbrowser_tests/tests/subgraph/subgraphNavigation.spec.tsbrowser_tests/tests/subgraph/subgraphNested.spec.tsbrowser_tests/tests/subgraph/subgraphOperations.spec.tsbrowser_tests/tests/subgraph/subgraphPromotion.spec.tsbrowser_tests/tests/subgraph/subgraphPromotionDom.spec.tsbrowser_tests/tests/subgraph/subgraphSearch.spec.tsbrowser_tests/tests/subgraph/subgraphSerialization.spec.tsbrowser_tests/tests/subgraph/subgraphSlots.spec.tsbrowser_tests/tests/subgraphInputSlotRename.spec.tsbrowser_tests/tests/subgraphLegacyPrefixedProxyWidgets.spec.tsbrowser_tests/tests/subgraphLifecycle.spec.tsbrowser_tests/tests/subgraphNestedConfigureOrder.spec.tsbrowser_tests/tests/subgraphNestedDuplicateWidgetNames.spec.tsbrowser_tests/tests/subgraphNestedPackValues.spec.tsbrowser_tests/tests/subgraphNestedStaleProxyWidgets.spec.tsbrowser_tests/tests/subgraphProgressClear.spec.tsbrowser_tests/tests/subgraphPromotion.spec.tsbrowser_tests/tests/subgraphSlotAlignment.spec.tsbrowser_tests/tests/subgraphViewport.spec.tssrc/lib/litegraph/src/subgraph/SubgraphNode.test.tssrc/lib/litegraph/src/subgraph/SubgraphSerialization.test.tssrc/lib/litegraph/src/subgraph/SubgraphWidgetPromotion.test.tssrc/stores/subgraphNavigationStore.test.ts
💤 Files with no reviewable changes (16)
- browser_tests/tests/subgraphNestedStaleProxyWidgets.spec.ts
- browser_tests/tests/subgraph-rename-dialog.spec.ts
- browser_tests/tests/subgraphLegacyPrefixedProxyWidgets.spec.ts
- browser_tests/tests/subgraphInputSlotRename.spec.ts
- browser_tests/tests/subgraph-promoted-widget-dom.spec.ts
- browser_tests/tests/subgraphSlotAlignment.spec.ts
- browser_tests/tests/subgraphProgressClear.spec.ts
- browser_tests/tests/subgraphNestedDuplicateWidgetNames.spec.ts
- browser_tests/tests/subgraph-promoted-slot-position.spec.ts
- browser_tests/tests/subgraphNestedPackValues.spec.ts
- browser_tests/tests/subgraphLifecycle.spec.ts
- browser_tests/tests/subgraphNestedConfigureOrder.spec.ts
- browser_tests/tests/subgraphViewport.spec.ts
- browser_tests/tests/subgraph-duplicate-ids.spec.ts
- browser_tests/tests/subgraphPromotion.spec.ts
- browser_tests/tests/subgraph.spec.ts
…'s code-quality patterns Amp-Thread-ID: https://ampcode.com/threads/T-019d407d-54fe-7159-a60d-22aa548d145a Co-authored-by: Amp <amp@ampcode.com>
…'s code-quality patterns Amp-Thread-ID: https://ampcode.com/threads/T-019d407d-54fe-7159-a60d-22aa548d145a Co-authored-by: Amp <amp@ampcode.com>
641ee1e to
5e56b26
Compare
🎭 Playwright: ✅ 979 passed, 0 failed · 2 flaky📊 Browser Reports
|
🎨 Storybook: ✅ Built — View Storybook |
📦 Bundle: 5.12 MB gzip 🔴 +214 BDetailsSummary
Category Glance App Entry Points — 22.3 kB (baseline 22.3 kB) • ⚪ 0 BMain entry bundles and manifests
Status: 1 added / 1 removed Graph Workspace — 1.2 MB (baseline 1.2 MB) • ⚪ 0 BGraph editor runtime, canvas, workflow orchestration
Status: 1 added / 1 removed Views & Navigation — 76.6 kB (baseline 76.6 kB) • ⚪ 0 BTop-level views, pages, and routed surfaces
Status: 9 added / 9 removed / 2 unchanged Panels & Settings — 484 kB (baseline 484 kB) • ⚪ 0 BConfiguration panels, inspectors, and settings screens
Status: 10 added / 10 removed / 12 unchanged User & Accounts — 17.1 kB (baseline 17.1 kB) • ⚪ 0 BAuthentication, profile, and account management bundles
Status: 5 added / 5 removed / 2 unchanged Editors & Dialogs — 109 kB (baseline 109 kB) • ⚪ 0 BModals, dialogs, drawers, and in-app editors
Status: 2 added / 2 removed UI Components — 60.3 kB (baseline 60.3 kB) • ⚪ 0 BReusable component library chunks
Status: 5 added / 5 removed / 8 unchanged Data & Services — 2.98 MB (baseline 2.98 MB) • ⚪ 0 BStores, services, APIs, and repositories
Status: 13 added / 13 removed / 4 unchanged Utilities & Hooks — 338 kB (baseline 338 kB) • ⚪ 0 BHelpers, composables, and utility bundles
Status: 13 added / 13 removed / 13 unchanged Vendor & Third-Party — 9.8 MB (baseline 9.8 MB) • ⚪ 0 BExternal libraries and shared vendor chunks Status: 16 unchanged Other — 8.53 MB (baseline 8.53 MB) • ⚪ 0 BBundles that do not match a named category
Status: 55 added / 55 removed / 79 unchanged ⚡ Performance Report
No regressions detected. All metrics
Historical variance (last 15 runs)
Trend (last 15 commits on main)
Raw data{
"timestamp": "2026-04-08T19:08:30.103Z",
"gitSha": "0f21958395b5dc04d93e8e50386643874650ff53",
"branch": "drjkl/test/subgraphs-3",
"measurements": [
{
"name": "canvas-idle",
"durationMs": 2010.126999999983,
"styleRecalcs": 12,
"styleRecalcDurationMs": 10.402999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 369.08799999999997,
"heapDeltaBytes": 20408900,
"heapUsedBytes": 63223364,
"domNodes": 23,
"jsHeapTotalBytes": 23330816,
"scriptDurationMs": 23.426000000000002,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-idle",
"durationMs": 2035.1229999999987,
"styleRecalcs": 11,
"styleRecalcDurationMs": 9.916,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 349.111,
"heapDeltaBytes": 20359664,
"heapUsedBytes": 64239948,
"domNodes": 22,
"jsHeapTotalBytes": 22806528,
"scriptDurationMs": 22.214,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "canvas-idle",
"durationMs": 2038.0349999999225,
"styleRecalcs": 10,
"styleRecalcDurationMs": 9.225999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 364.939,
"heapDeltaBytes": 20398060,
"heapUsedBytes": 63197868,
"domNodes": 20,
"jsHeapTotalBytes": 23068672,
"scriptDurationMs": 22.559999999999995,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1989.7890000000018,
"styleRecalcs": 79,
"styleRecalcDurationMs": 46.004999999999995,
"layouts": 12,
"layoutDurationMs": 4.1370000000000005,
"taskDurationMs": 858.1200000000001,
"heapDeltaBytes": 17911648,
"heapUsedBytes": 66991092,
"domNodes": 63,
"jsHeapTotalBytes": 25690112,
"scriptDurationMs": 148.624,
"eventListeners": 4,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1870.636999999988,
"styleRecalcs": 77,
"styleRecalcDurationMs": 39.425000000000004,
"layouts": 12,
"layoutDurationMs": 3.8369999999999997,
"taskDurationMs": 767.816,
"heapDeltaBytes": 16670476,
"heapUsedBytes": 59259788,
"domNodes": 62,
"jsHeapTotalBytes": 23330816,
"scriptDurationMs": 134.75799999999998,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1840.8809999999676,
"styleRecalcs": 75,
"styleRecalcDurationMs": 38.355,
"layouts": 12,
"layoutDurationMs": 3.5769999999999995,
"taskDurationMs": 748.908,
"heapDeltaBytes": 16723920,
"heapUsedBytes": 59444364,
"domNodes": 58,
"jsHeapTotalBytes": 23855104,
"scriptDurationMs": 136.411,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1732.9369999999926,
"styleRecalcs": 32,
"styleRecalcDurationMs": 17.906,
"layouts": 6,
"layoutDurationMs": 0.525,
"taskDurationMs": 294.526,
"heapDeltaBytes": 24774592,
"heapUsedBytes": 67241324,
"domNodes": 80,
"jsHeapTotalBytes": 20709376,
"scriptDurationMs": 24.394,
"eventListeners": 19,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1746.6000000000008,
"styleRecalcs": 31,
"styleRecalcDurationMs": 18.068,
"layouts": 6,
"layoutDurationMs": 1.4500000000000002,
"taskDurationMs": 304.53799999999995,
"heapDeltaBytes": 24633020,
"heapUsedBytes": 67499340,
"domNodes": 77,
"jsHeapTotalBytes": 20447232,
"scriptDurationMs": 25.039,
"eventListeners": 19,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1733.3750000000236,
"styleRecalcs": 32,
"styleRecalcDurationMs": 17.44,
"layouts": 6,
"layoutDurationMs": 0.5219999999999999,
"taskDurationMs": 306.819,
"heapDeltaBytes": 24981428,
"heapUsedBytes": 68809512,
"domNodes": 79,
"jsHeapTotalBytes": 20709376,
"scriptDurationMs": 25.210000000000004,
"eventListeners": 19,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "dom-widget-clipping",
"durationMs": 580.2489999999807,
"styleRecalcs": 13,
"styleRecalcDurationMs": 10.488,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 368.348,
"heapDeltaBytes": 6534656,
"heapUsedBytes": 49337804,
"domNodes": 22,
"jsHeapTotalBytes": 13107200,
"scriptDurationMs": 74.922,
"eventListeners": 2,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000273
},
{
"name": "dom-widget-clipping",
"durationMs": 572.6480000000151,
"styleRecalcs": 13,
"styleRecalcDurationMs": 9.508,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 335.089,
"heapDeltaBytes": 6845904,
"heapUsedBytes": 49332756,
"domNodes": 22,
"jsHeapTotalBytes": 13107200,
"scriptDurationMs": 66.57499999999999,
"eventListeners": 2,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000273
},
{
"name": "dom-widget-clipping",
"durationMs": 539.3159999999853,
"styleRecalcs": 13,
"styleRecalcDurationMs": 9.126999999999997,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 340.15999999999997,
"heapDeltaBytes": -2703368,
"heapUsedBytes": 49148476,
"domNodes": 22,
"jsHeapTotalBytes": 15728640,
"scriptDurationMs": 64.10300000000001,
"eventListeners": 2,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999727
},
{
"name": "large-graph-idle",
"durationMs": 2053.078999999997,
"styleRecalcs": 11,
"styleRecalcDurationMs": 11.061000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 554.948,
"heapDeltaBytes": 5921720,
"heapUsedBytes": 56311780,
"domNodes": -257,
"jsHeapTotalBytes": 16969728,
"scriptDurationMs": 99.413,
"eventListeners": -125,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-idle",
"durationMs": 2023.3479999999986,
"styleRecalcs": 11,
"styleRecalcDurationMs": 10.194999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 534.181,
"heapDeltaBytes": 4005992,
"heapUsedBytes": 54952488,
"domNodes": -257,
"jsHeapTotalBytes": 16707584,
"scriptDurationMs": 95.056,
"eventListeners": -125,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-idle",
"durationMs": 2055.8320000000094,
"styleRecalcs": 10,
"styleRecalcDurationMs": 10.301000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 546.4610000000001,
"heapDeltaBytes": 5208296,
"heapUsedBytes": 55894100,
"domNodes": -257,
"jsHeapTotalBytes": 16707584,
"scriptDurationMs": 104.535,
"eventListeners": -127,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-pan",
"durationMs": 2145.0649999999882,
"styleRecalcs": 69,
"styleRecalcDurationMs": 18.436,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1078.762,
"heapDeltaBytes": 16812356,
"heapUsedBytes": 68746384,
"domNodes": -260,
"jsHeapTotalBytes": 17960960,
"scriptDurationMs": 396.292,
"eventListeners": -127,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-pan",
"durationMs": 2134.9820000000364,
"styleRecalcs": 70,
"styleRecalcDurationMs": 17.090000000000003,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1076.157,
"heapDeltaBytes": 16963300,
"heapUsedBytes": 70327960,
"domNodes": -260,
"jsHeapTotalBytes": 18223104,
"scriptDurationMs": 385.608,
"eventListeners": -127,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.670000000000012,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-pan",
"durationMs": 2138.516000000095,
"styleRecalcs": 68,
"styleRecalcDurationMs": 15.214999999999996,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1090.251,
"heapDeltaBytes": 15766316,
"heapUsedBytes": 68864520,
"domNodes": -261,
"jsHeapTotalBytes": 18223104,
"scriptDurationMs": 399.031,
"eventListeners": -127,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-zoom",
"durationMs": 3164.274000000006,
"styleRecalcs": 66,
"styleRecalcDurationMs": 17.265,
"layouts": 60,
"layoutDurationMs": 7.434999999999999,
"taskDurationMs": 1310.791,
"heapDeltaBytes": 7443084,
"heapUsedBytes": 62192596,
"domNodes": -261,
"jsHeapTotalBytes": 17756160,
"scriptDurationMs": 487.792,
"eventListeners": -123,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-zoom",
"durationMs": 3164.445999999998,
"styleRecalcs": 66,
"styleRecalcDurationMs": 16.57,
"layouts": 60,
"layoutDurationMs": 7.458999999999999,
"taskDurationMs": 1312.4920000000002,
"heapDeltaBytes": 9290236,
"heapUsedBytes": 64259216,
"domNodes": -263,
"jsHeapTotalBytes": 15921152,
"scriptDurationMs": 493.20000000000005,
"eventListeners": -123,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-zoom",
"durationMs": 3173.1079999999565,
"styleRecalcs": 67,
"styleRecalcDurationMs": 24.721999999999998,
"layouts": 60,
"layoutDurationMs": 7.407,
"taskDurationMs": 1349.672,
"heapDeltaBytes": 7719088,
"heapUsedBytes": 62132768,
"domNodes": -261,
"jsHeapTotalBytes": 16969728,
"scriptDurationMs": 498.02,
"eventListeners": -123,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "minimap-idle",
"durationMs": 2018.4169999999995,
"styleRecalcs": 9,
"styleRecalcDurationMs": 7.854,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 538.1009999999999,
"heapDeltaBytes": 3097056,
"heapUsedBytes": 56820636,
"domNodes": -261,
"jsHeapTotalBytes": 15921152,
"scriptDurationMs": 94.55199999999999,
"eventListeners": -127,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "minimap-idle",
"durationMs": 2009.5059999999876,
"styleRecalcs": 9,
"styleRecalcDurationMs": 9.179,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 541.3509999999999,
"heapDeltaBytes": 3972060,
"heapUsedBytes": 56509900,
"domNodes": -260,
"jsHeapTotalBytes": 15921152,
"scriptDurationMs": 97.986,
"eventListeners": -127,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "minimap-idle",
"durationMs": 2054.8929999999928,
"styleRecalcs": 10,
"styleRecalcDurationMs": 10.248,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 547.099,
"heapDeltaBytes": 17772280,
"heapUsedBytes": 71317976,
"domNodes": -260,
"jsHeapTotalBytes": 15454208,
"scriptDurationMs": 112.822,
"eventListeners": -129,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 543.5809999999606,
"styleRecalcs": 47,
"styleRecalcDurationMs": 11.366,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 358.783,
"heapDeltaBytes": 6447144,
"heapUsedBytes": 49596552,
"domNodes": 20,
"jsHeapTotalBytes": 13369344,
"scriptDurationMs": 124.84899999999999,
"eventListeners": 8,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000273
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 559.1009999999983,
"styleRecalcs": 48,
"styleRecalcDurationMs": 17.597,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 365.563,
"heapDeltaBytes": 6437556,
"heapUsedBytes": 49542228,
"domNodes": 22,
"jsHeapTotalBytes": 14155776,
"scriptDurationMs": 129.26500000000001,
"eventListeners": 8,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000273
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 592.2379999999521,
"styleRecalcs": 48,
"styleRecalcDurationMs": 12.822,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 376.97,
"heapDeltaBytes": 6524040,
"heapUsedBytes": 49602408,
"domNodes": 22,
"jsHeapTotalBytes": 13107200,
"scriptDurationMs": 130.876,
"eventListeners": 8,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-idle",
"durationMs": 2012.4289999999974,
"styleRecalcs": 10,
"styleRecalcDurationMs": 9.423000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 344.472,
"heapDeltaBytes": 19989608,
"heapUsedBytes": 63030264,
"domNodes": 20,
"jsHeapTotalBytes": 22806528,
"scriptDurationMs": 17.964000000000002,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-idle",
"durationMs": 1995.317,
"styleRecalcs": 11,
"styleRecalcDurationMs": 9.685,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 340.26,
"heapDeltaBytes": 19886640,
"heapUsedBytes": 63101296,
"domNodes": 22,
"jsHeapTotalBytes": 23068672,
"scriptDurationMs": 18.371,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "subgraph-idle",
"durationMs": 1993.5880000000452,
"styleRecalcs": 11,
"styleRecalcDurationMs": 10.636,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 354.06899999999996,
"heapDeltaBytes": 19982084,
"heapUsedBytes": 63014308,
"domNodes": 22,
"jsHeapTotalBytes": 22544384,
"scriptDurationMs": 19.969,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1987.1029999999905,
"styleRecalcs": 84,
"styleRecalcDurationMs": 47.503,
"layouts": 16,
"layoutDurationMs": 4.998,
"taskDurationMs": 942.271,
"heapDeltaBytes": 12270384,
"heapUsedBytes": 55290608,
"domNodes": 73,
"jsHeapTotalBytes": 23330816,
"scriptDurationMs": 105.088,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1733.5009999999897,
"styleRecalcs": 76,
"styleRecalcDurationMs": 36.31099999999999,
"layouts": 16,
"layoutDurationMs": 4.8870000000000005,
"taskDurationMs": 664.02,
"heapDeltaBytes": 11865828,
"heapUsedBytes": 55000212,
"domNodes": 63,
"jsHeapTotalBytes": 22806528,
"scriptDurationMs": 94.278,
"eventListeners": 4,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1968.195000000037,
"styleRecalcs": 83,
"styleRecalcDurationMs": 45.547999999999995,
"layouts": 16,
"layoutDurationMs": 4.810999999999999,
"taskDurationMs": 903.389,
"heapDeltaBytes": 12304892,
"heapUsedBytes": 55384412,
"domNodes": 71,
"jsHeapTotalBytes": 23330816,
"scriptDurationMs": 102.328,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "viewport-pan-sweep",
"durationMs": 8224.38199999999,
"styleRecalcs": 253,
"styleRecalcDurationMs": 45.698,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 3692.0769999999993,
"heapDeltaBytes": 28552200,
"heapUsedBytes": 79100260,
"domNodes": -253,
"jsHeapTotalBytes": 19271680,
"scriptDurationMs": 1267.334,
"eventListeners": -109,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "viewport-pan-sweep",
"durationMs": 8150.000000000035,
"styleRecalcs": 251,
"styleRecalcDurationMs": 44.115,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 3642.1079999999997,
"heapDeltaBytes": 25516384,
"heapUsedBytes": 76011860,
"domNodes": -255,
"jsHeapTotalBytes": 19795968,
"scriptDurationMs": 1232.499,
"eventListeners": -111,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "viewport-pan-sweep",
"durationMs": 8183.532000000014,
"styleRecalcs": 250,
"styleRecalcDurationMs": 42.349,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 3838.278,
"heapDeltaBytes": 26675760,
"heapUsedBytes": 77955892,
"domNodes": -260,
"jsHeapTotalBytes": 20320256,
"scriptDurationMs": 1320.817,
"eventListeners": -111,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-idle",
"durationMs": 12166.839999999978,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 12155.333000000002,
"heapDeltaBytes": -50625624,
"heapUsedBytes": 145947948,
"domNodes": -8331,
"jsHeapTotalBytes": 18440192,
"scriptDurationMs": 615.9870000000001,
"eventListeners": -16462,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.223333333333358,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-idle",
"durationMs": 12233.78500000001,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 12222.536000000002,
"heapDeltaBytes": -37769788,
"heapUsedBytes": 157413716,
"domNodes": -8331,
"jsHeapTotalBytes": 16867328,
"scriptDurationMs": 636.5039999999999,
"eventListeners": -16462,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.220000000000073,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-idle",
"durationMs": 12194.438999999988,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 12182.029,
"heapDeltaBytes": -44473464,
"heapUsedBytes": 166266800,
"domNodes": -8333,
"jsHeapTotalBytes": 28925952,
"scriptDurationMs": 591.9380000000001,
"eventListeners": -16462,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "vue-large-graph-pan",
"durationMs": 14427.180999999962,
"styleRecalcs": 66,
"styleRecalcDurationMs": 13.647999999999993,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 14400.895999999999,
"heapDeltaBytes": -34957360,
"heapUsedBytes": 172071260,
"domNodes": -8333,
"jsHeapTotalBytes": -1306624,
"scriptDurationMs": 859.221,
"eventListeners": -16456,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.776666666666763,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-pan",
"durationMs": 13907.106999999996,
"styleRecalcs": 64,
"styleRecalcDurationMs": 13.613000000000042,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 13888.829000000002,
"heapDeltaBytes": -26118120,
"heapUsedBytes": 173355164,
"domNodes": -8331,
"jsHeapTotalBytes": 25432064,
"scriptDurationMs": 868.512,
"eventListeners": -16460,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.220000000000073,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-pan",
"durationMs": 14498.688000000016,
"styleRecalcs": 70,
"styleRecalcDurationMs": 14.468000000000009,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 14477.547999999999,
"heapDeltaBytes": -52241032,
"heapUsedBytes": 148453296,
"domNodes": -8331,
"jsHeapTotalBytes": -696320,
"scriptDurationMs": 888.7159999999999,
"eventListeners": -16488,
"totalBlockingTimeMs": 35,
"frameDurationMs": 17.223333333333358,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "workflow-execution",
"durationMs": 452.87799999999834,
"styleRecalcs": 17,
"styleRecalcDurationMs": 24.505999999999997,
"layouts": 5,
"layoutDurationMs": 1.3840000000000001,
"taskDurationMs": 128.38899999999998,
"heapDeltaBytes": 4743900,
"heapUsedBytes": 49529728,
"domNodes": 167,
"jsHeapTotalBytes": 262144,
"scriptDurationMs": 30.723,
"eventListeners": 71,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000273
},
{
"name": "workflow-execution",
"durationMs": 467.3440000000255,
"styleRecalcs": 23,
"styleRecalcDurationMs": 27.790000000000003,
"layouts": 5,
"layoutDurationMs": 1.735,
"taskDurationMs": 129.492,
"heapDeltaBytes": 4523776,
"heapUsedBytes": 50106756,
"domNodes": 183,
"jsHeapTotalBytes": 262144,
"scriptDurationMs": 25.964000000000002,
"eventListeners": 71,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "workflow-execution",
"durationMs": 460.4669999999942,
"styleRecalcs": 16,
"styleRecalcDurationMs": 20.743000000000002,
"layouts": 5,
"layoutDurationMs": 1.271,
"taskDurationMs": 117.93800000000002,
"heapDeltaBytes": 4493596,
"heapUsedBytes": 49262636,
"domNodes": 156,
"jsHeapTotalBytes": 262144,
"scriptDurationMs": 28.858999999999995,
"eventListeners": 71,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.663333333333338,
"p95FrameDurationMs": 16.800000000000182
}
]
} |
There was a problem hiding this comment.
Actionable comments posted: 1
♻️ Duplicate comments (2)
browser_tests/tests/subgraph/subgraphPromotion.spec.ts (1)
161-200:⚠️ Potential issue | 🟡 MinorSelect the node before opening widget context menus in both promote/un-promote paths.
Both right-click flows should select the KSampler node first to keep context-menu targeting deterministic.
Suggested patch
const ksampler = await comfyPage.nodeOps.getNodeRefById('1') + await ksampler.click('title') const stepsWidget = await ksampler.getWidget(2) const widgetPos = await stepsWidget.getPosition() @@ const subgraphNode2 = await comfyPage.nodeOps.getNodeRefById('2') await subgraphNode2.navigateIntoSubgraph() - const stepsWidget2 = await ( - await comfyPage.nodeOps.getNodeRefById('1') - ).getWidget(2) + const ksampler2 = await comfyPage.nodeOps.getNodeRefById('1') + await ksampler2.click('title') + const stepsWidget2 = await ksampler2.getWidget(2) const widgetPos2 = await stepsWidget2.getPosition()As per coding guidelines: Select a node first using
vueNodes.selectNode()ornodeRef.click('title')before interacting with context menus to ensure menu items appear correctly.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphPromotion.spec.ts` around lines 161 - 200, The context-menu right-click flows don’t select the target node first, causing nondeterministic menu targeting; before calling comfyPage.canvas.click({... button: 'right' ...}) in both promote and un-promote paths, call the node selection helper (e.g., await vueNodes.selectNode(nodeRef) or await nodeRef.click('title')) on the node returned by comfyPage.nodeOps.getNodeRefById('1') (and similarly for node '2' when navigating into the subgraph) so the KSampler node is explicitly selected prior to opening the widget context menu.browser_tests/tests/subgraph/subgraphLifecycle.spec.ts (1)
51-64:⚠️ Potential issue | 🟡 MinorAssert preview DOM presence before deletion to avoid false-positive cleanup passes.
The final
.toHaveCount(0)can still pass if the preview DOM never rendered. Add a precondition before deleting node5so this test actually validates cleanup.Suggested patch
const beforePseudo = await getPseudoPreviewWidgets(comfyPage, '5') expect(beforePseudo.length).toBeGreaterThan(0) + await expect(comfyPage.page.locator(domPreviewSelector)).toHaveCount(1) const subgraphNode = await comfyPage.nodeOps.getNodeRefById('5') expect(await subgraphNode.exists()).toBe(true)🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphLifecycle.spec.ts` around lines 51 - 64, Before deleting node '5', assert the preview DOM is actually rendered by adding a precondition that the locator identified by domPreviewSelector has at least one element; e.g., after obtaining beforePseudo (getPseudoPreviewWidgets) and before calling subgraphNode.delete(), add an assertion using comfyPage.page.locator(domPreviewSelector) (or await locator.count()) to verify count > 0 so the subsequent cleanup assertions cannot pass spuriously when the preview never rendered.
🧹 Nitpick comments (1)
browser_tests/tests/subgraph/subgraphNavigation.spec.ts (1)
88-90: Inconsistent selector pattern: avoid chaining raw CSS after TestId.Lines 52-54 use
getByTestId(TestIds.breadcrumb.subgraph)directly, but here you chain.locator('.p-breadcrumb'). This couples the test to PrimeVue's internal DOM structure and creates inconsistency within the same file.If the intent is to scope to the PrimeVue breadcrumb component specifically, consider either extending TestIds or using the same direct TestId pattern used elsewhere in this file.
♻️ Suggested simplification
- const breadcrumb = comfyPage.page - .getByTestId(TestIds.breadcrumb.subgraph) - .locator('.p-breadcrumb') + const breadcrumb = comfyPage.page.getByTestId(TestIds.breadcrumb.subgraph)🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphNavigation.spec.ts` around lines 88 - 90, The test is coupling to PrimeVue internals by chaining .locator('.p-breadcrumb') after comfyPage.page.getByTestId(TestIds.breadcrumb.subgraph); instead, remove the raw CSS chain and either use the existing TestId directly (comfyPage.page.getByTestId(TestIds.breadcrumb.subgraph)) or add a new TestIds entry for the specific breadcrumb element and use that TestId consistently (update TestIds and replace the locator call), so tests stop depending on '.p-breadcrumb' DOM structure.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts`:
- Around line 385-399: The test is still locating the widget by the old literal
label 'seed' after renaming to RENAMED_LABEL; update the UI assertion to look
for the renamed text by replacing getByLabel('seed', { exact: true }) with a
call that uses the RENAMED_LABEL constant (e.g.,
subgraphNodeAfter.getByLabel(RENAMED_LABEL, { exact: true })) and ensure
subsequent checks (seedWidgetAfter and SubgraphHelper.expectWidgetBelowHeader)
operate on that renamed locator so the spec validates visible rename
propagation.
---
Duplicate comments:
In `@browser_tests/tests/subgraph/subgraphLifecycle.spec.ts`:
- Around line 51-64: Before deleting node '5', assert the preview DOM is
actually rendered by adding a precondition that the locator identified by
domPreviewSelector has at least one element; e.g., after obtaining beforePseudo
(getPseudoPreviewWidgets) and before calling subgraphNode.delete(), add an
assertion using comfyPage.page.locator(domPreviewSelector) (or await
locator.count()) to verify count > 0 so the subsequent cleanup assertions cannot
pass spuriously when the preview never rendered.
In `@browser_tests/tests/subgraph/subgraphPromotion.spec.ts`:
- Around line 161-200: The context-menu right-click flows don’t select the
target node first, causing nondeterministic menu targeting; before calling
comfyPage.canvas.click({... button: 'right' ...}) in both promote and un-promote
paths, call the node selection helper (e.g., await vueNodes.selectNode(nodeRef)
or await nodeRef.click('title')) on the node returned by
comfyPage.nodeOps.getNodeRefById('1') (and similarly for node '2' when
navigating into the subgraph) so the KSampler node is explicitly selected prior
to opening the widget context menu.
---
Nitpick comments:
In `@browser_tests/tests/subgraph/subgraphNavigation.spec.ts`:
- Around line 88-90: The test is coupling to PrimeVue internals by chaining
.locator('.p-breadcrumb') after
comfyPage.page.getByTestId(TestIds.breadcrumb.subgraph); instead, remove the raw
CSS chain and either use the existing TestId directly
(comfyPage.page.getByTestId(TestIds.breadcrumb.subgraph)) or add a new TestIds
entry for the specific breadcrumb element and use that TestId consistently
(update TestIds and replace the locator call), so tests stop depending on
'.p-breadcrumb' DOM structure.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: cc4eb030-6619-4e7c-a3c7-4481b0e2c354
📒 Files selected for processing (5)
browser_tests/tests/subgraph/subgraphLifecycle.spec.tsbrowser_tests/tests/subgraph/subgraphNavigation.spec.tsbrowser_tests/tests/subgraph/subgraphNested.spec.tsbrowser_tests/tests/subgraph/subgraphPromotion.spec.tsbrowser_tests/tests/subgraph/subgraphSlots.spec.ts
🚧 Files skipped from review as they are similar to previous changes (1)
- browser_tests/tests/subgraph/subgraphNested.spec.ts
Amp-Thread-ID: https://ampcode.com/threads/T-019d40a2-847a-77ac-bb10-165b70958fb6 Co-authored-by: Amp <amp@ampcode.com>
There was a problem hiding this comment.
♻️ Duplicate comments (2)
browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts (1)
109-120:⚠️ Potential issue | 🟡 MinorUse retrying locator assertions for teardown count checks.
Line 109 and Line 117 use snapshot
count(); this can flake afterdelete()when cleanup lands on the next frame. Please switch totoHaveCount(...).As per coding guidelines: `browser_tests/**/*.spec.ts`: Follow Playwright best practices documented in Playwright documentation.Suggested patch
- const initialCount = await comfyPage.page - .locator(DOM_WIDGET_SELECTOR) - .count() - expect(initialCount).toBe(1) + const textareas = comfyPage.page.locator(DOM_WIDGET_SELECTOR) + await expect(textareas).toHaveCount(1) @@ - const finalCount = await comfyPage.page - .locator(DOM_WIDGET_SELECTOR) - .count() - expect(finalCount).toBe(0) + await expect(textareas).toHaveCount(0)🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts` around lines 109 - 120, The snapshot-style .count() assertions are flaky around teardown; replace the two count() + expect(...) checks with Playwright's retrying locator assertions by calling the locator for DOM_WIDGET_SELECTOR and using toHaveCount(1) for the initial expected state and toHaveCount(0) after calling subgraphNode.delete(); update the assertions that reference comfyPage.page.locator(DOM_WIDGET_SELECTOR) and the sequence around subgraphNode.delete() to use await locator.toHaveCount(...) so the check retries until the DOM stabilizes.browser_tests/tests/subgraph/subgraphLifecycle.spec.ts (1)
51-53:⚠️ Potential issue | 🟡 MinorAdd a DOM-present precondition before the cleanup action.
Line 51 and Line 76 only prove internal proxy state exists. If preview DOM never renders, Line 65/Line 89 can still pass. Please assert
.image-previewis present before unpack/delete so this truly exercises DOM cleanup.Suggested patch
const beforePseudo = await getPseudoPreviewWidgets(comfyPage, '5') expect(beforePseudo.length).toBeGreaterThan(0) + await expect(comfyPage.page.locator(domPreviewSelector)).toHaveCount(1) @@ const beforePseudo = await getPseudoPreviewWidgets(comfyPage, '5') expect(beforePseudo.length).toBeGreaterThan(0) + await expect(comfyPage.page.locator(domPreviewSelector)).toHaveCount(1)Also applies to: 76-77
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphLifecycle.spec.ts` around lines 51 - 53, The test currently only checks internal proxy state via getPseudoPreviewWidgets(comfyPage, '5') but may never verify the preview DOM rendered before performing cleanup; add an explicit DOM-present precondition by asserting the '.image-preview' element exists (using comfyPage helpers like waitForSelector or querying and asserting non-null) immediately before each cleanup/unpack/delete action (the spots around getPseudoPreviewWidgets calls at the two locations), so the test truly exercises DOM cleanup rather than just internal state.
🧹 Nitpick comments (1)
browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts (1)
132-136: Prefer retryingtoHaveCountfor render/navigation count assertions.Line 132 and Line 158+ use snapshot counts during async view transitions. Converting these to retrying assertions will reduce intermittent flakes and align with the rest of the file’s pattern.
As per coding guidelines:
browser_tests/**/*.spec.ts: Follow Playwright best practices documented in Playwright documentation.Also applies to: 158-178
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts` around lines 132 - 136, The test is using an immediate snapshot count check (const textareaCount = await comfyPage.page.locator(DOM_WIDGET_SELECTOR).count(); expect(textareaCount).toBe(1)) which can flake during async navigation; replace these with Playwright's retrying assertion by calling expect(comfyPage.page.locator(DOM_WIDGET_SELECTOR)).toHaveCount(1) (await the expect) for both occurrences (the one using textareaCount and the similar count assertion later in the file) so the assertion retries until the expected count appears.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Duplicate comments:
In `@browser_tests/tests/subgraph/subgraphLifecycle.spec.ts`:
- Around line 51-53: The test currently only checks internal proxy state via
getPseudoPreviewWidgets(comfyPage, '5') but may never verify the preview DOM
rendered before performing cleanup; add an explicit DOM-present precondition by
asserting the '.image-preview' element exists (using comfyPage helpers like
waitForSelector or querying and asserting non-null) immediately before each
cleanup/unpack/delete action (the spots around getPseudoPreviewWidgets calls at
the two locations), so the test truly exercises DOM cleanup rather than just
internal state.
In `@browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts`:
- Around line 109-120: The snapshot-style .count() assertions are flaky around
teardown; replace the two count() + expect(...) checks with Playwright's
retrying locator assertions by calling the locator for DOM_WIDGET_SELECTOR and
using toHaveCount(1) for the initial expected state and toHaveCount(0) after
calling subgraphNode.delete(); update the assertions that reference
comfyPage.page.locator(DOM_WIDGET_SELECTOR) and the sequence around
subgraphNode.delete() to use await locator.toHaveCount(...) so the check retries
until the DOM stabilizes.
---
Nitpick comments:
In `@browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts`:
- Around line 132-136: The test is using an immediate snapshot count check
(const textareaCount = await
comfyPage.page.locator(DOM_WIDGET_SELECTOR).count();
expect(textareaCount).toBe(1)) which can flake during async navigation; replace
these with Playwright's retrying assertion by calling
expect(comfyPage.page.locator(DOM_WIDGET_SELECTOR)).toHaveCount(1) (await the
expect) for both occurrences (the one using textareaCount and the similar count
assertion later in the file) so the assertion retries until the expected count
appears.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: b37c900b-5ac6-499f-b9f2-ea9ade3c1bda
📒 Files selected for processing (2)
browser_tests/tests/subgraph/subgraphLifecycle.spec.tsbrowser_tests/tests/subgraph/subgraphPromotionDom.spec.ts
- Tighten queue response assertion to response.ok() (subgraphNested) - Use ControlOrMeta for cross-platform copy/paste (subgraphOperations) - Use RENAMED_LABEL in post-rename locator assertion (subgraphSlots) - Wrap Alt key drag in try/finally to prevent leaks (subgraphCrud) - Add comment explaining double nextFrame intent (subgraphSlots) Amp-Thread-ID: https://ampcode.com/threads/T-019d40d8-0fd2-72b2-9422-607306079abd Co-authored-by: Amp <amp@ampcode.com>
There was a problem hiding this comment.
🧹 Nitpick comments (1)
browser_tests/tests/subgraph/subgraphSlots.spec.ts (1)
476-480: Clarify thewidgetNameassertion intent.The assertion
expect(after!.widgetName).not.toBe('my_custom_prompt')verifies that the underlying widget name is unchanged after renaming the slot label. However, the assertion reads as if it's checking that something is not the new name, which could be confusing. Consider adding a brief comment or asserting whatwidgetNameshould be (the original widget name) for clarity.💡 Suggested improvement
expect(after!.hasPos).toBe(true) expect(after!.posY).toBeGreaterThan(after!.titleHeight) + // Widget name should remain unchanged; only the slot label was renamed expect(after!.widgetName).not.toBe('my_custom_prompt')Or assert the positive case if the original widget name is known:
expect(after!.widgetName).toBe(before!.widgetName)🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts` around lines 476 - 480, The negative assertion on after!.widgetName is ambiguous; change it to assert the positive expected value to make intent clear. Replace expect(after!.widgetName).not.toBe('my_custom_prompt') with an explicit comparison to the original value (e.g., expect(after!.widgetName).toBe(before!.widgetName)) or add a one-line comment above the assertion explaining that widgetName must remain unchanged; update the test in subgraphSlots.spec.ts around the after!/before! checks accordingly.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts`:
- Around line 476-480: The negative assertion on after!.widgetName is ambiguous;
change it to assert the positive expected value to make intent clear. Replace
expect(after!.widgetName).not.toBe('my_custom_prompt') with an explicit
comparison to the original value (e.g.,
expect(after!.widgetName).toBe(before!.widgetName)) or add a one-line comment
above the assertion explaining that widgetName must remain unchanged; update the
test in subgraphSlots.spec.ts around the after!/before! checks accordingly.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 7ae03c1a-6789-45f7-85f0-551d343f4e0d
📒 Files selected for processing (4)
browser_tests/tests/subgraph/subgraphCrud.spec.tsbrowser_tests/tests/subgraph/subgraphNested.spec.tsbrowser_tests/tests/subgraph/subgraphOperations.spec.tsbrowser_tests/tests/subgraph/subgraphSlots.spec.ts
✅ Files skipped from review due to trivial changes (1)
- browser_tests/tests/subgraph/subgraphNested.spec.ts
The widget's aria-label stays 'seed' after renaming; assert the visible label text 'my_seed' separately instead of querying by the renamed aria-label. Amp-Thread-ID: https://ampcode.com/threads/T-019d410e-20b9-735c-a6d3-d703ec83951f Co-authored-by: Amp <amp@ampcode.com>
There was a problem hiding this comment.
Comprehensive PR Review
This review is generated by Claude. It may not always be accurate, as with human reviewers. If you believe that any of the comments are invalid or incorrect, please state why for each. For others, please implement the changes in one way or another.
Review Summary
PR: test: reorganize subgraph test suite into composable domain specs (#10759)
Impact: 791 additions, 1985 deletions across 14 files
Type: Test reorganization and refinement
Issue Distribution
- Critical: 0
- High: 0
- Medium: 1
- Low: 3
Category Breakdown
- Architecture: 1 issue (positive observation)
- Security: 0 issues
- Performance: 0 issues
- Code Quality: 3 issues
Key Findings
Test Organization & Architecture
Strengths:
- Excellent domain-driven test organization: splitting 96 scattered browser tests into 54 focused user journeys with clear ownership boundaries
- Strong coverage migration from Playwright to Vitest for semantic/combinatorial cases, following the Practical Test Pyramid model
- Clear separation of concerns: browser tests own UI-only failure modes, unit tests own data semantics
- Proper use of try-finally for cleanup in keyboard modifier operations (subgraphCrud.spec.ts:21-28)
- Enhanced event verification in SubgraphNode unit tests ensures event-driven architecture reliability
Observations:
- Test count reduction (96 → 54 browser tests) with semantic coverage preserved in 230 Vitest tests demonstrates mature test pyramid thinking
- Domain-specific organization (slots, promotion, navigation, lifecycle, CRUD) makes test ownership and maintenance clearer
- The PR description explicitly documents coverage ownership split, which is excellent documentation practice
Code Quality & Maintainability
Minor improvements identified:
- Platform compatibility: Some keyboard shortcuts could use cross-platform helpers (e.g., Control+a vs ControlOrMeta+a)
- Timeout documentation: 20-second timeout in navigation tests may benefit from investigation or documentation of why such a long wait is needed
- API validation depth: API response checks use .ok() but could validate specific status codes for stronger contract testing
- Comment precision: Some inline comments (e.g., "Two frames needed...") would benefit from more specificity about what happens in each frame
Testing Best Practices Applied
Excellent patterns observed:
- Defensive cleanup with try-finally blocks for modifier keys
- Use of TestIds instead of fragile CSS selectors
- Proper use of expect.poll() for async state assertions
- Console warning collection in SubgraphHelper for regression prevention
- Event spy verification for graph trigger events
Integration & Extension Impact
No breaking changes: This is a pure test refactoring with no product code changes. All test reorganization maintains behavioral coverage while improving organization and reducing redundancy.
Positive Observations
- Test suite maturity: The migration demonstrates a mature understanding of the test pyramid - browser tests handle UI journeys, Vitest handles business logic
- Clear ownership model: Each test now has a single owner and clear responsibility
- Improved maintainability: Domain-organized specs will be easier to navigate and maintain than the previous flat structure
- Strong defensive coding: Try-finally cleanup patterns and event verification prevent flaky tests
- Documentation quality: PR description clearly explains the reorganization strategy and coverage ownership
Next Steps
- Address the 4 minor quality observations in inline comments (none are blocking)
- Consider extracting keyboard modifier cleanup into a reusable helper for consistency across the test suite
- Monitor CI performance to ensure the test reorganization maintains or improves execution time
- Consider adding similar event verification tests for other critical state mutations (as suggested in inline comment)
This is a comprehensive automated review. The test reorganization is well-executed and follows software engineering best practices. No critical issues found - this PR is in excellent shape for merge.
There was a problem hiding this comment.
🧹 Nitpick comments (1)
browser_tests/tests/subgraph/subgraphSlots.spec.ts (1)
480-484: Clarify assertion intent: verify widgetName is unchanged rather than just not matching the new label.The negative assertion
expect(after!.widgetName).not.toBe('my_custom_prompt')only verifies the widget name isn't the new label value, but doesn't confirm what it actually is. If the intent is to verify that renaming the slot label doesn't affect the widget's internal name, comparing before/after would be more robust.Proposed fix for clearer assertion
expect(after!.hasPos).toBe(true) expect(after!.posY).toBeGreaterThan(after!.titleHeight) - expect(after!.widgetName).not.toBe('my_custom_prompt') + expect(after!.widgetName).toBe(before!.widgetName) })🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts` around lines 480 - 484, The assertion checking slot widget identity is too weak—replace the negative check on after!.widgetName with a direct comparison to the original value (compare before!.widgetName to after!.widgetName) so the test confirms the widgetName did not change when the slot label was renamed; locate the test block that uses variables before and after (and the assertions on after!.hasPos, after!.posY, after!.titleHeight) and update the final expect to assert equality between before!.widgetName and after!.widgetName.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts`:
- Around line 480-484: The assertion checking slot widget identity is too
weak—replace the negative check on after!.widgetName with a direct comparison to
the original value (compare before!.widgetName to after!.widgetName) so the test
confirms the widgetName did not change when the slot label was renamed; locate
the test block that uses variables before and after (and the assertions on
after!.hasPos, after!.posY, after!.titleHeight) and update the final expect to
assert equality between before!.widgetName and after!.widgetName.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 6490c3c9-c96d-4bb3-9885-afe48e52a9e4
📒 Files selected for processing (1)
browser_tests/tests/subgraph/subgraphSlots.spec.ts
- Select node before widget context menu right-click (subgraphPromotion) - Use retrying toHaveCount assertion (subgraphPromotionDom) - Isolate dummy node-type registration in beforeEach/afterEach (SubgraphSerialization) Amp-Thread-ID: https://ampcode.com/threads/T-019d4195-1a01-77d9-b200-dd8f1f667519 Co-authored-by: Amp <amp@ampcode.com>
- Remove .p-breadcrumb PrimeVue coupling in navigation test - Use navigateIntoSubgraph() instead of canvas.setGraph() bypass - Add DOM precondition assertions before cleanup in lifecycle tests - Convert snapshot .count() to retrying toHaveCount() assertions - Use positive widgetName assertion after slot rename - Document double nextFrame intent Amp-Thread-ID: https://ampcode.com/threads/T-019d4c3f-d285-7680-9aa2-ad41cdf1930c Co-authored-by: Amp <amp@ampcode.com>
…visibility Amp-Thread-ID: https://ampcode.com/threads/T-019d6901-c741-70fb-a5d8-8a20c9e3de26 Co-authored-by: Amp <amp@ampcode.com>
- Add clone coverage test for SubgraphNode proxyWidgets preservation - Un-export PromotedWidgetEntry (zero external consumers) - Fix vi.resetAllMocks() ordering in subgraphNavigationStore tests - Replace as ComfyWorkflow casts with fromPartial<ComfyWorkflow>() Amp-Thread-ID: https://ampcode.com/threads/T-019d698b-746d-76b8-afda-1b02db3e7953 Co-authored-by: Amp <amp@ampcode.com>
Amp-Thread-ID: https://ampcode.com/threads/T-019d6e22-8206-75f8-9211-dbab497d2430 Co-authored-by: Amp <amp@ampcode.com> # Conflicts: # browser_tests/tests/subgraph/subgraphNavigation.spec.ts # browser_tests/tests/subgraph/subgraphPromotion.spec.ts # browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts # browser_tests/tests/subgraph/subgraphSerialization.spec.ts # browser_tests/tests/subgraph/subgraphSlots.spec.ts
Amp-Thread-ID: https://ampcode.com/threads/T-019d6e22-8206-75f8-9211-dbab497d2430 Co-authored-by: Amp <amp@ampcode.com> # Conflicts: # browser_tests/tests/subgraph/subgraphCrud.spec.ts # browser_tests/tests/subgraph/subgraphLifecycle.spec.ts # browser_tests/tests/subgraph/subgraphPromotionDom.spec.ts # browser_tests/tests/subgraph/subgraphSerialization.spec.ts # browser_tests/tests/subgraph/subgraphSlots.spec.ts
There was a problem hiding this comment.
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
browser_tests/tests/subgraph/subgraphPromotion.spec.ts (2)
63-78:⚠️ Potential issue | 🟡 MinorTest intent and assertions diverge in the pseudo-widget promotion case.
The title says “regular and pseudo-widget promotions,” but the test only asserts
'text'. Add an explicit pseudo-widget assertion (e.g., a promoted name starting with$$).Suggested patch
const nodeId = String(subgraphNode.id) // CLIPTextEncode is in the recommendedNodes list, so its text widget // should be promoted await expectPromotedWidgetNamesToContain(comfyPage, nodeId, 'text') + await expect + .poll( + async () => + (await getPromotedWidgetNames(comfyPage, nodeId)).some((name) => + name.startsWith('$$') + ), + { timeout: 5000 } + ) + .toBe(true)🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphPromotion.spec.ts` around lines 63 - 78, The test "Preview-capable nodes keep regular and pseudo-widget promotions when converted" only asserts the regular 'text' widget promotion; add an explicit assertion for the pseudo-widget promotion after converting the CLIPTextEncode node to a subgraph by calling expectPromotedWidgetNamesToContain with the same nodeId but checking for a promoted name that starts with "$$" (pseudo-widget naming convention). Locate the test block using the test title, the calls to comfyPage.nodeOps.getNodeRefById('6') and clipNode.convertToSubgraph(), and add the pseudo-widget assertion right after the existing expectPromotedWidgetNamesToContain(comfyPage, nodeId, 'text') call so both regular and pseudo-widget promotions are verified.
381-417:⚠️ Potential issue | 🟠 Major
Vue Mode - Promoted Preview Contentcurrently has no runnable test.This describe block only has
test.beforeEach(...); without atest(...), none of these assertions execute.Suggested patch
test.describe('Vue Mode - Promoted Preview Content', () => { - test.beforeEach(async ({ comfyPage }) => { + test('renders promoted preview content without pseudo-widget placeholder text', async ({ + comfyPage + }) => { await comfyPage.settings.setSetting('Comfy.VueNodes.Enabled', true) await comfyPage.workflow.loadWorkflow( 'subgraphs/subgraph-with-preview-node' @@ '$$canvas-image-preview' ) }) })🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphPromotion.spec.ts` around lines 381 - 417, The describe block "Vue Mode - Promoted Preview Content" contains only test.beforeEach and therefore no runnable test; move the assertions into a real test by adding a test(...) inside the test.describe (or convert the beforeEach to a test) so the assertions execute: keep the setup in test.beforeEach (Comfy.VueNodes.Enabled, workflow load, waitForNodes, dragDrop and command.executeCommand) and create a test named e.g. 'shows promoted preview content' that performs the checks using getPromotedWidgetNames(comfyPage, '5'), subgraphVueNode locator '[data-testid="node-body-5"]', expectations on .lg-node-widgets and .image-preview img count, and the negative text check for '$$canvas-image-preview'.
♻️ Duplicate comments (1)
browser_tests/tests/subgraph/subgraphLifecycle.spec.ts (1)
51-53:⚠️ Potential issue | 🟡 MinorAdd a DOM precondition before asserting cleanup to zero.
Both tests verify final
.image-previewcount is0, but neither proves the preview DOM existed before unpack/delete. That allows false passes on render regressions.Suggested patch
const beforePseudo = await getPseudoPreviewWidgets(comfyPage, '5') expect(beforePseudo.length).toBeGreaterThan(0) + await expect(comfyPage.page.locator(domPreviewSelector)).toHaveCount(1) @@ const beforePseudo = await getPseudoPreviewWidgets(comfyPage, '5') expect(beforePseudo.length).toBeGreaterThan(0) + await expect(comfyPage.page.locator(domPreviewSelector)).toHaveCount(1)Also applies to: 76-89
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphLifecycle.spec.ts` around lines 51 - 53, Add a DOM precondition asserting the preview existed before asserting cleanup: call getPseudoPreviewWidgets(comfyPage, '5') (as in the existing beforePseudo usage) immediately before the unpack/delete action and assert its length is > 0, then proceed with the action and assert final .image-preview count is 0; repeat the same precondition insertion for the other test block that covers lines 76-89 so both tests prove the preview DOM existed prior to cleanup.
🧹 Nitpick comments (1)
browser_tests/tests/subgraph/subgraphSlots.spec.ts (1)
146-148: Consider usingexpect(locator).toBeVisible()for consistency.This file mixes
page.waitForSelector()(lines 146, 171, 202) withexpect(locator).toBeVisible()(lines 124, 263, 297, etc.) for dialog visibility checks. The latter is the modern Playwright pattern with auto-retrying semantics.♻️ Suggested consistency fix
- await comfyPage.page.waitForSelector(SELECTORS.promptDialog, { - state: 'visible' - }) + await expect(comfyPage.page.locator(SELECTORS.promptDialog)).toBeVisible()Apply similar changes at lines 171-173 and 202-204.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts` around lines 146 - 148, The test uses page.waitForSelector for dialog visibility (e.g., comfyPage.page.waitForSelector(SELECTORS.promptDialog)) while other checks use Playwright's expect(locator).toBeVisible(); replace the waitForSelector calls with the corresponding locator and use await expect(locator).toBeVisible() to get consistent auto-retry semantics — locate uses of comfyPage.page.waitForSelector(SELECTORS.promptDialog) and similar calls at the noted spots and change them to await expect(comfyPage.page.locator(SELECTORS.promptDialog)).toBeVisible(), applying the same pattern for the other occurrences mentioned.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Outside diff comments:
In `@browser_tests/tests/subgraph/subgraphPromotion.spec.ts`:
- Around line 63-78: The test "Preview-capable nodes keep regular and
pseudo-widget promotions when converted" only asserts the regular 'text' widget
promotion; add an explicit assertion for the pseudo-widget promotion after
converting the CLIPTextEncode node to a subgraph by calling
expectPromotedWidgetNamesToContain with the same nodeId but checking for a
promoted name that starts with "$$" (pseudo-widget naming convention). Locate
the test block using the test title, the calls to
comfyPage.nodeOps.getNodeRefById('6') and clipNode.convertToSubgraph(), and add
the pseudo-widget assertion right after the existing
expectPromotedWidgetNamesToContain(comfyPage, nodeId, 'text') call so both
regular and pseudo-widget promotions are verified.
- Around line 381-417: The describe block "Vue Mode - Promoted Preview Content"
contains only test.beforeEach and therefore no runnable test; move the
assertions into a real test by adding a test(...) inside the test.describe (or
convert the beforeEach to a test) so the assertions execute: keep the setup in
test.beforeEach (Comfy.VueNodes.Enabled, workflow load, waitForNodes, dragDrop
and command.executeCommand) and create a test named e.g. 'shows promoted preview
content' that performs the checks using getPromotedWidgetNames(comfyPage, '5'),
subgraphVueNode locator '[data-testid="node-body-5"]', expectations on
.lg-node-widgets and .image-preview img count, and the negative text check for
'$$canvas-image-preview'.
---
Duplicate comments:
In `@browser_tests/tests/subgraph/subgraphLifecycle.spec.ts`:
- Around line 51-53: Add a DOM precondition asserting the preview existed before
asserting cleanup: call getPseudoPreviewWidgets(comfyPage, '5') (as in the
existing beforePseudo usage) immediately before the unpack/delete action and
assert its length is > 0, then proceed with the action and assert final
.image-preview count is 0; repeat the same precondition insertion for the other
test block that covers lines 76-89 so both tests prove the preview DOM existed
prior to cleanup.
---
Nitpick comments:
In `@browser_tests/tests/subgraph/subgraphSlots.spec.ts`:
- Around line 146-148: The test uses page.waitForSelector for dialog visibility
(e.g., comfyPage.page.waitForSelector(SELECTORS.promptDialog)) while other
checks use Playwright's expect(locator).toBeVisible(); replace the
waitForSelector calls with the corresponding locator and use await
expect(locator).toBeVisible() to get consistent auto-retry semantics — locate
uses of comfyPage.page.waitForSelector(SELECTORS.promptDialog) and similar calls
at the noted spots and change them to await
expect(comfyPage.page.locator(SELECTORS.promptDialog)).toBeVisible(), applying
the same pattern for the other occurrences mentioned.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: c211619e-e65f-4390-892c-8ee4a9c8a36c
📒 Files selected for processing (11)
browser_tests/helpers/promotedWidgets.tsbrowser_tests/tests/subgraph/subgraphCrud.spec.tsbrowser_tests/tests/subgraph/subgraphLifecycle.spec.tsbrowser_tests/tests/subgraph/subgraphNavigation.spec.tsbrowser_tests/tests/subgraph/subgraphNested.spec.tsbrowser_tests/tests/subgraph/subgraphOperations.spec.tsbrowser_tests/tests/subgraph/subgraphPromotion.spec.tsbrowser_tests/tests/subgraph/subgraphPromotionDom.spec.tsbrowser_tests/tests/subgraph/subgraphSearch.spec.tsbrowser_tests/tests/subgraph/subgraphSerialization.spec.tsbrowser_tests/tests/subgraph/subgraphSlots.spec.ts
✅ Files skipped from review due to trivial changes (1)
- browser_tests/tests/subgraph/subgraphNavigation.spec.ts
🚧 Files skipped from review as they are similar to previous changes (2)
- browser_tests/tests/subgraph/subgraphSearch.spec.ts
- browser_tests/tests/subgraph/subgraphOperations.spec.ts
Codecov Report✅ All modified and coverable lines are covered by tests. @@ Coverage Diff @@
## main #10759 +/- ##
========================================
Coverage 43.08% 43.08%
========================================
Files 1332 1332
Lines 68165 68165
Branches 18812 17955 -857
========================================
Hits 29366 29366
Misses 38203 38203
Partials 596 596 🚀 New features to boost your workflow:
|
*PR Created by the Glary-Bot Agent* --- ## Summary Comfy-Org#10759 removed ~12 E2E tests from `subgraphSerialization.spec.ts` during a reorganization that shifted semantic coverage to Vitest. Several of the removed tests covered the exact `serialize() → JSON → configure()` round-trip that Comfy-Org#10849's positional `_instanceWidgetValues` path later regressed on Main — promoted widget values binding to the wrong slots when loading templates whose `widgets_values` ordering doesn't match current `proxyWidgets`. This restores the pre-reorg E2E coverage so future regressions in promoted-widget serialization are caught at the browser level. ## Restored tests From `subgraphSerialization.spec.ts` (pre-Comfy-Org#10759): - **Deterministic proxyWidgets Hydrate** (3 tests) — round-trip stability and compressed `target_slot` resolution. - **Legacy And Round-Trip Coverage** (5 tests) — includes the most directly-relevant restorations: - `Promoted widgets survive serialize -> loadGraphData round-trip` - `Multi-link input representative stays stable through save/reload` - `Cloning a subgraph node keeps promoted widget entries on original and clone` - **Duplicate ID Remapping** (5 tests) — includes `Promoted widget tuples are stable after full page reload boot path`. The 4 tests that already existed on Main (added by Comfy-Org#10849 and the Vue-nodes legacy-prefixed block) are kept as-is. ## Adaptations to current APIs - Imports reworked for the post-`@e2e/*` alias layout. - Redundant `comfyPage.nextFrame()` calls dropped — `loadWorkflow` / `loadGraphData` / `serializeAndReload` already wait internally (Comfy-Org#11264). - Alt-drag clone block wrapped in `try/finally` around `keyboard.up('Alt')` to match the current `subgraphCrud.spec.ts` pattern. - `PromotedWidgetEntry` is now exported from `browser_tests/helpers/promotedWidgets.ts` so the restored `expectPromotedWidgetsToResolveToInteriorNodes` helper can type its argument. ## Review follow-ups applied - Use `expect.poll()` instead of `expect(async () => …).toPass()` for the single-value snapshot comparison, per `browser_tests/AGENTS.md`. - Capture and call `dispose()` from `SubgraphHelper.collectConsoleWarnings()` inside a `try/finally` so the console listener is unregistered after the test. ## Verification - `pnpm typecheck:browser` — clean. - `pnpm exec eslint` + `pnpm exec oxlint` on changed files — 0 warnings, 0 errors. - `pnpm exec oxfmt` on changed files — applied (no diff). - Ran 2 key restored tests against local ComfyUI + dev server: - `Promoted widgets survive serialize -> loadGraphData round-trip` — PASS (3.9s) - `Multi-link input representative stays stable through save/reload` — PASS (2.9s) - Full-suite runs in this sandbox are blocked by the existing `comfyPage` fixture's `createUser` path failing on repeat runs against persistent backend state — unrelated to this PR. CI will exercise the full suite. ┆Issue is synchronized with this [Notion page](https://www.notion.so/PR-11579-test-restore-deleted-subgraph-serialization-E2E-tests-34b6d73d365081f29b27c1069476ad17) by [Unito](https://www.unito.io) --------- Co-authored-by: Glary-Bot <glary-bot@users.noreply.github.com>
…0759) ## Summary Reorganize the subgraph test suite so browser tests are thin representative user journeys while lower-level Vitest suites own combinatorics, migration edge cases, and data-shape semantics. ## Changes - **What**: Migrate 17 flat subgraph browser specs into 10 domain-organized specs under `browser_tests/tests/subgraph/`, move redundant semantic coverage down to 8 Vitest owner suites, delete all legacy flat files - **Browser specs** (54 tests): `subgraphSlots`, `subgraphPromotion`, `subgraphPromotionDom`, `subgraphSerialization`, `subgraphNavigation`, `subgraphNested`, `subgraphLifecycle`, `subgraphCrud`, `subgraphSearch`, `subgraphOperations` - **Vitest owners** (230 tests): `SubgraphNode.test.ts` (rename/label propagation), `subgraphNodePromotion.test.ts`, `promotedWidgetView.test.ts`, `SubgraphSerialization.test.ts` (duplicate-ID remap), `SubgraphWidgetPromotion.test.ts` (legacy hydration), `subgraphNavigationStore*.test.ts` (viewport cache, workflow-switch), `subgraphStore.test.ts` (search aliases, description) - **Net effect**: browser suite shrinks from ~96 scattered tests to 54 focused journeys ## Review Focus - Coverage ownership split: each browser test has a unique UI-only failure mode; semantic coverage lives in Vitest - `subgraphPromotionDom.spec.ts` forces LiteGraph mode and uses `canvas.openSubgraph()` instead of `navigateIntoSubgraph()` to avoid a wrapper-specific DOM overlay duplication issue — entry-affordance coverage lives in `subgraphNavigation.spec.ts` - No product code changes — test-only migration ┆Issue is synchronized with this [Notion page](https://www.notion.so/PR-10759-test-reorganize-subgraph-test-suite-into-composable-domain-specs-3336d73d365081b0a56bcbf809b1f584) by [Unito](https://www.unito.io) --------- Co-authored-by: Amp <amp@ampcode.com>
*PR Created by the Glary-Bot Agent* --- ## Summary #10759 removed ~12 E2E tests from `subgraphSerialization.spec.ts` during a reorganization that shifted semantic coverage to Vitest. Several of the removed tests covered the exact `serialize() → JSON → configure()` round-trip that #10849's positional `_instanceWidgetValues` path later regressed on Main — promoted widget values binding to the wrong slots when loading templates whose `widgets_values` ordering doesn't match current `proxyWidgets`. This restores the pre-reorg E2E coverage so future regressions in promoted-widget serialization are caught at the browser level. ## Restored tests From `subgraphSerialization.spec.ts` (pre-#10759): - **Deterministic proxyWidgets Hydrate** (3 tests) — round-trip stability and compressed `target_slot` resolution. - **Legacy And Round-Trip Coverage** (5 tests) — includes the most directly-relevant restorations: - `Promoted widgets survive serialize -> loadGraphData round-trip` - `Multi-link input representative stays stable through save/reload` - `Cloning a subgraph node keeps promoted widget entries on original and clone` - **Duplicate ID Remapping** (5 tests) — includes `Promoted widget tuples are stable after full page reload boot path`. The 4 tests that already existed on Main (added by #10849 and the Vue-nodes legacy-prefixed block) are kept as-is. ## Adaptations to current APIs - Imports reworked for the post-`@e2e/*` alias layout. - Redundant `comfyPage.nextFrame()` calls dropped — `loadWorkflow` / `loadGraphData` / `serializeAndReload` already wait internally (#11264). - Alt-drag clone block wrapped in `try/finally` around `keyboard.up('Alt')` to match the current `subgraphCrud.spec.ts` pattern. - `PromotedWidgetEntry` is now exported from `browser_tests/helpers/promotedWidgets.ts` so the restored `expectPromotedWidgetsToResolveToInteriorNodes` helper can type its argument. ## Review follow-ups applied - Use `expect.poll()` instead of `expect(async () => …).toPass()` for the single-value snapshot comparison, per `browser_tests/AGENTS.md`. - Capture and call `dispose()` from `SubgraphHelper.collectConsoleWarnings()` inside a `try/finally` so the console listener is unregistered after the test. ## Verification - `pnpm typecheck:browser` — clean. - `pnpm exec eslint` + `pnpm exec oxlint` on changed files — 0 warnings, 0 errors. - `pnpm exec oxfmt` on changed files — applied (no diff). - Ran 2 key restored tests against local ComfyUI + dev server: - `Promoted widgets survive serialize -> loadGraphData round-trip` — PASS (3.9s) - `Multi-link input representative stays stable through save/reload` — PASS (2.9s) - Full-suite runs in this sandbox are blocked by the existing `comfyPage` fixture's `createUser` path failing on repeat runs against persistent backend state — unrelated to this PR. CI will exercise the full suite. ┆Issue is synchronized with this [Notion page](https://www.notion.so/PR-11579-test-restore-deleted-subgraph-serialization-E2E-tests-34b6d73d365081f29b27c1069476ad17) by [Unito](https://www.unito.io) --------- Co-authored-by: Glary-Bot <glary-bot@users.noreply.github.com>
Summary
Reorganize the subgraph test suite so browser tests are thin representative user journeys while lower-level Vitest suites own combinatorics, migration edge cases, and data-shape semantics.
Changes
browser_tests/tests/subgraph/, move redundant semantic coverage down to 8 Vitest owner suites, delete all legacy flat filessubgraphSlots,subgraphPromotion,subgraphPromotionDom,subgraphSerialization,subgraphNavigation,subgraphNested,subgraphLifecycle,subgraphCrud,subgraphSearch,subgraphOperationsSubgraphNode.test.ts(rename/label propagation),subgraphNodePromotion.test.ts,promotedWidgetView.test.ts,SubgraphSerialization.test.ts(duplicate-ID remap),SubgraphWidgetPromotion.test.ts(legacy hydration),subgraphNavigationStore*.test.ts(viewport cache, workflow-switch),subgraphStore.test.ts(search aliases, description)Review Focus
subgraphPromotionDom.spec.tsforces LiteGraph mode and usescanvas.openSubgraph()instead ofnavigateIntoSubgraph()to avoid a wrapper-specific DOM overlay duplication issue — entry-affordance coverage lives insubgraphNavigation.spec.ts┆Issue is synchronized with this Notion page by Unito