Repository navigation
fix: autogrow slot can't be reconnected after disconnect without refresh (PM-1496) - #18289
Conversation
|
Important Review skippedBot user detected. To trigger a single review, invoke the ⚙️ Run configurationConfiguration used: Repository: Comfy-Org/ComfyUI_frontend/.coderabbit.yaml Review profile: ASSERTIVE Plan: Advanced Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Comment |
🎭 Playwright: ✅ 2076 passed, 0 failed · 1 flaky📊 Browser Reports
🎨 Storybook: ✅ Built — View Storybook🌾 Fallow: ✅ PassedFallow audits the diff for duplication, complexity and dead code. It gates on new findings only, so when the audit completes, a red is something this PR introduced or moved — never inherited backlog. A red can also mean the audit itself did not finish; the details below say which happened. No new findings in the changed files. Reproduce locally# the base the gate compared against
npx fallow audit --root . --quiet --format json \
--changed-since 2a3d472b6d0fdfdb7ee842b217b2d85c42ba5f2c
Full docs: ADR-DEVEX-LINT-0015 🧩 Custom Nodes Ecosystem Matrix: ✅ PassedExecutes the frontend JS of every registry pack that ships any, against this change, in both renderers. The eight
On a pin-bump PR, a red gate means the ecosystem moved — not that your diff broke something. That is what the bump PR is for. Outside a pin bump or a cold rebuild of the unpinned tail, a red means the diff. Note the corpus size and the executed count are different numbers, and the verdict is against the executed count: roughly 3,100 of ~5,100 packs ship no extension-shaped JS and are skipped. Reproduce locally# build one shard's fixture, then run it
python3 scripts/registry-census/build_matrix.py --shard 1/4
MATRIX_RENDERER=vue MATRIX_VUE=1 \
npx vitest run --config vitest.matrix.config.mts
Full docs: scripts/registry-census/README.md 📦 Bundle: 9.96 MB gzip 🟢 -28 BDetailsSummary
Category Glance App Entry Points — 3.87 kB (baseline 3.87 kB) • ⚪ 0 BMain entry bundles and manifests
Status: 1 added / 1 removed Graph Workspace — 1.46 MB (baseline 1.46 MB) • ⚪ 0 BGraph editor runtime, canvas, workflow orchestration
Status: 2 added / 2 removed / 1 unchanged Views & Navigation — 132 kB (baseline 132 kB) • ⚪ 0 BTop-level views, pages, and routed surfaces
Status: 14 added / 14 removed / 4 unchanged Panels & Settings — 639 kB (baseline 639 kB) • ⚪ 0 BConfiguration panels, inspectors, and settings screens
Status: 10 added / 10 removed / 18 unchanged User & Accounts — 37.9 kB (baseline 37.9 kB) • ⚪ 0 BAuthentication, profile, and account management bundles
Status: 6 added / 6 removed / 5 unchanged Editors & Dialogs — 126 kB (baseline 126 kB) • ⚪ 0 BModals, dialogs, drawers, and in-app editors
Status: 7 added / 7 removed / 1 unchanged UI Components — 123 kB (baseline 123 kB) • ⚪ 0 BReusable component library chunks
Status: 7 added / 7 removed / 9 unchanged Data & Services — 3.88 MB (baseline 3.88 MB) • 🔴 +11 BStores, services, APIs, and repositories
Status: 13 added / 13 removed / 3 unchanged Utilities & Hooks — 577 kB (baseline 577 kB) • ⚪ 0 BHelpers, composables, and utility bundles
Status: 19 added / 19 removed / 19 unchanged Vendor & Third-Party — 17.4 MB (baseline 17.4 MB) • ⚪ 0 BExternal libraries and shared vendor chunks Status: 18 unchanged Other — 17.8 MB (baseline 17.8 MB) • ⚪ 0 BBundles that do not match a named category
Status: 74 added / 74 removed / 228 unchanged ⚡ Performance Report
Show regressions
All metrics
Historical variance (last 15 runs)
Trend (last 15 commits on main)
Raw data{
"timestamp": "2026-09-22T04:37:34.854Z",
"gitSha": "699198bb0baf1db0ca8aec5085804c133d1964c7",
"branch": "test/autogrow-disconnect-pending-connection-race",
"measurements": [
{
"name": "canvas-idle",
"durationMs": 1993.678999999986,
"styleRecalcs": 9,
"styleRecalcDurationMs": 6.587000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 307.956,
"heapDeltaBytes": 16631368,
"heapUsedBytes": 78684988,
"domNodes": -258,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 9.004,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-idle",
"durationMs": 1989.6580000000768,
"styleRecalcs": 10,
"styleRecalcDurationMs": 8.747,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 337.716,
"heapDeltaBytes": -9298156,
"heapUsedBytes": 58008936,
"domNodes": -260,
"jsHeapTotalBytes": 4972544,
"scriptDurationMs": 8.272,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1692.1909999999798,
"styleRecalcs": 73,
"styleRecalcDurationMs": 24.908999999999995,
"layouts": 14,
"layoutDurationMs": 4.634,
"taskDurationMs": 602.317,
"heapDeltaBytes": -3390512,
"heapUsedBytes": 63889928,
"domNodes": -258,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 115.91300000000001,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1677.0320000000538,
"styleRecalcs": 73,
"styleRecalcDurationMs": 27.058000000000003,
"layouts": 14,
"layoutDurationMs": 4.659,
"taskDurationMs": 664.044,
"heapDeltaBytes": 18461908,
"heapUsedBytes": 80503804,
"domNodes": -263,
"jsHeapTotalBytes": 4972544,
"scriptDurationMs": 122.041,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1721.1540000000127,
"styleRecalcs": 33,
"styleRecalcDurationMs": 15.748999999999999,
"layouts": 6,
"layoutDurationMs": 1.346,
"taskDurationMs": 280.015,
"heapDeltaBytes": 2826228,
"heapUsedBytes": 70182556,
"domNodes": 79,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 14.750000000000002,
"eventListeners": 15,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1719.549000000029,
"styleRecalcs": 31,
"styleRecalcDurationMs": 16.004,
"layouts": 6,
"layoutDurationMs": 1.2540000000000002,
"taskDurationMs": 264.175,
"heapDeltaBytes": 2858140,
"heapUsedBytes": 70038780,
"domNodes": 78,
"jsHeapTotalBytes": 5505024,
"scriptDurationMs": 12.954,
"eventListeners": 15,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-clipping",
"durationMs": 513.9749999999594,
"styleRecalcs": 11,
"styleRecalcDurationMs": 6.939000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 319.39300000000003,
"heapDeltaBytes": 18513160,
"heapUsedBytes": 80532940,
"domNodes": 18,
"jsHeapTotalBytes": 5767168,
"scriptDurationMs": 115.122,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.663333333333338,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "dom-widget-clipping",
"durationMs": 430.7659999999487,
"styleRecalcs": 12,
"styleRecalcDurationMs": 7.3229999999999995,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 297.587,
"heapDeltaBytes": 18184268,
"heapUsedBytes": 80408500,
"domNodes": 20,
"jsHeapTotalBytes": 5767168,
"scriptDurationMs": 106.961,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000273
},
{
"name": "dom-widget-order-dense-n200-w200-i120",
"durationMs": 886.1879999999474,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 528.8050000000001,
"heapDeltaBytes": -17758116,
"heapUsedBytes": 72200660,
"domNodes": -295,
"jsHeapTotalBytes": 21364736,
"scriptDurationMs": 37.35999999999999,
"eventListeners": -217,
"totalBlockingTimeMs": 266,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "dom-widget-order-dense-n200-w200-i120",
"durationMs": 936.8390000000772,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 593.148,
"heapDeltaBytes": -13087724,
"heapUsedBytes": 72883492,
"domNodes": -297,
"jsHeapTotalBytes": 19791872,
"scriptDurationMs": 42.66299999999999,
"eventListeners": -217,
"totalBlockingTimeMs": 307,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "dom-widget-order-sparse-n1000-w20-i120",
"durationMs": 686.6980000000353,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 271.0379999999999,
"heapDeltaBytes": 14239120,
"heapUsedBytes": 89171704,
"domNodes": 0,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 93.93299999999999,
"eventListeners": 0,
"totalBlockingTimeMs": 91,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-order-sparse-n1000-w20-i120",
"durationMs": 696.6709999999239,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 303.4789999999998,
"heapDeltaBytes": 15304820,
"heapUsedBytes": 91615268,
"domNodes": 0,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 101.73199999999999,
"eventListeners": 0,
"totalBlockingTimeMs": 94,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "extension-clean-combined-draw",
"durationMs": 6553.7180000000035,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 5493.2080000000005,
"heapDeltaBytes": 12604632,
"heapUsedBytes": 94172824,
"domNodes": -16,
"jsHeapTotalBytes": 9175040,
"scriptDurationMs": 4272.101000000001,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "extension-clean-combined-draw",
"durationMs": 6404.189999999971,
"styleRecalcs": 1,
"styleRecalcDurationMs": 0.7149999999999969,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 5434.806,
"heapDeltaBytes": -10582464,
"heapUsedBytes": 71645608,
"domNodes": -16,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 4195.898999999999,
"eventListeners": 0,
"totalBlockingTimeMs": 43,
"frameDurationMs": 16.66333333333338,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "extension-rgthree-prototype-draw",
"durationMs": 6565.373999999963,
"styleRecalcs": 1,
"styleRecalcDurationMs": 0.688000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 5474.916,
"heapDeltaBytes": 2337784,
"heapUsedBytes": 84538336,
"domNodes": -16,
"jsHeapTotalBytes": 7077888,
"scriptDurationMs": 4286.071,
"eventListeners": 0,
"totalBlockingTimeMs": 46,
"frameDurationMs": 17.220000000000013,
"p95FrameDurationMs": 16.80000000000109
},
{
"name": "extension-rgthree-prototype-draw",
"durationMs": 6505.018000000064,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 5611.968,
"heapDeltaBytes": 6961608,
"heapUsedBytes": 80431988,
"domNodes": -10,
"jsHeapTotalBytes": 3670016,
"scriptDurationMs": 4315.864,
"eventListeners": 0,
"totalBlockingTimeMs": 25,
"frameDurationMs": 17.220000000000013,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "large-graph-idle",
"durationMs": 2021.8760000000202,
"styleRecalcs": 7,
"styleRecalcDurationMs": 6.1560000000000015,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 475.03800000000007,
"heapDeltaBytes": 2315264,
"heapUsedBytes": 83418936,
"domNodes": -246,
"jsHeapTotalBytes": -1806336,
"scriptDurationMs": 32.451,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-idle",
"durationMs": 2023.3090000000402,
"styleRecalcs": 7,
"styleRecalcDurationMs": 8.909000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 464.09200000000004,
"heapDeltaBytes": 3493524,
"heapUsedBytes": 83515212,
"domNodes": -242,
"jsHeapTotalBytes": -794624,
"scriptDurationMs": 31.476999999999997,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-pan",
"durationMs": 2359.5449999999687,
"styleRecalcs": 68,
"styleRecalcDurationMs": 14.104000000000003,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1203.0789999999997,
"heapDeltaBytes": -1434428,
"heapUsedBytes": 80227768,
"domNodes": -246,
"jsHeapTotalBytes": 253952,
"scriptDurationMs": 585.083,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-pan",
"durationMs": 2121.7020000000275,
"styleRecalcs": 68,
"styleRecalcDurationMs": 13.578999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1096.0539999999999,
"heapDeltaBytes": -1960236,
"heapUsedBytes": 79809644,
"domNodes": -246,
"jsHeapTotalBytes": -270336,
"scriptDurationMs": 561.258,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-zoom",
"durationMs": 3136.179000000027,
"styleRecalcs": 62,
"styleRecalcDurationMs": 13.542000000000002,
"layouts": 60,
"layoutDurationMs": 15.667,
"taskDurationMs": 1253.598,
"heapDeltaBytes": 7996424,
"heapUsedBytes": 74107608,
"domNodes": 6,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 590.1469999999999,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-zoom",
"durationMs": 3122.967000000017,
"styleRecalcs": 63,
"styleRecalcDurationMs": 13.351000000000003,
"layouts": 60,
"layoutDurationMs": 14.474999999999998,
"taskDurationMs": 1182.111,
"heapDeltaBytes": 9871000,
"heapUsedBytes": 74796816,
"domNodes": 8,
"jsHeapTotalBytes": 524288,
"scriptDurationMs": 537.55,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "legacy-node-drag",
"durationMs": 2557.939000000033,
"styleRecalcs": 43,
"styleRecalcDurationMs": 8.024000000000003,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1464.298,
"heapDeltaBytes": 17944792,
"heapUsedBytes": 83142116,
"domNodes": 6,
"jsHeapTotalBytes": 417792,
"scriptDurationMs": 721.559,
"eventListeners": 180,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "legacy-node-drag",
"durationMs": 2319.205000000011,
"styleRecalcs": 45,
"styleRecalcDurationMs": 8.770999999999997,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1392.301,
"heapDeltaBytes": 18508972,
"heapUsedBytes": 82541120,
"domNodes": -251,
"jsHeapTotalBytes": 679936,
"scriptDurationMs": 726.206,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "minimap-idle",
"durationMs": 2013.0080000000135,
"styleRecalcs": 8,
"styleRecalcDurationMs": 6.5619999999999985,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 420.7860000000001,
"heapDeltaBytes": -4778896,
"heapUsedBytes": 84025788,
"domNodes": -242,
"jsHeapTotalBytes": -1282048,
"scriptDurationMs": 29.781000000000002,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "minimap-idle",
"durationMs": 2013.153999999986,
"styleRecalcs": 6,
"styleRecalcDurationMs": 4.407000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 369.4649999999999,
"heapDeltaBytes": -5396780,
"heapUsedBytes": 83709776,
"domNodes": -249,
"jsHeapTotalBytes": -1019904,
"scriptDurationMs": 28.960999999999995,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 456.8170000000009,
"styleRecalcs": 48,
"styleRecalcDurationMs": 11.118,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 291.725,
"heapDeltaBytes": 12801780,
"heapUsedBytes": 76522216,
"domNodes": 22,
"jsHeapTotalBytes": 6291456,
"scriptDurationMs": 101.19600000000001,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.663333333333338,
"p95FrameDurationMs": 16.700000000000273
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 412.45200000003024,
"styleRecalcs": 47,
"styleRecalcDurationMs": 10.143,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 275.541,
"heapDeltaBytes": 13712860,
"heapUsedBytes": 83405640,
"domNodes": 20,
"jsHeapTotalBytes": 5767168,
"scriptDurationMs": 97.61,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000273
},
{
"name": "subgraph-idle",
"durationMs": 1991.9340000000147,
"styleRecalcs": 11,
"styleRecalcDurationMs": 8.359,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 345.043,
"heapDeltaBytes": 11820176,
"heapUsedBytes": 81471276,
"domNodes": -275,
"jsHeapTotalBytes": 4972544,
"scriptDurationMs": 9.589,
"eventListeners": -198,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-idle",
"durationMs": 2005.145999999968,
"styleRecalcs": 11,
"styleRecalcDurationMs": 7.117999999999998,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 301.458,
"heapDeltaBytes": 17792604,
"heapUsedBytes": 81404648,
"domNodes": -273,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 6.08,
"eventListeners": -198,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1690.1429999999777,
"styleRecalcs": 79,
"styleRecalcDurationMs": 27.016,
"layouts": 18,
"layoutDurationMs": 5.799,
"taskDurationMs": 633.1419999999999,
"heapDeltaBytes": 439716,
"heapUsedBytes": 64096060,
"domNodes": -273,
"jsHeapTotalBytes": 6283264,
"scriptDurationMs": 96.283,
"eventListeners": -220,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1688.4590000000799,
"styleRecalcs": 79,
"styleRecalcDurationMs": 26.672,
"layouts": 18,
"layoutDurationMs": 5.590000000000001,
"taskDurationMs": 581.3999999999999,
"heapDeltaBytes": 16962644,
"heapUsedBytes": 80516008,
"domNodes": -274,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 87.798,
"eventListeners": -198,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "subgraph-transition-enter",
"durationMs": 833.9299999999525,
"styleRecalcs": 18,
"styleRecalcDurationMs": 21.236000000000004,
"layouts": 5,
"layoutDurationMs": 8.678999999999998,
"taskDurationMs": 663.376,
"heapDeltaBytes": 20159620,
"heapUsedBytes": 102622968,
"domNodes": 13667,
"jsHeapTotalBytes": 10747904,
"scriptDurationMs": 22.3,
"eventListeners": 2383,
"totalBlockingTimeMs": 77,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "viewport-pan-sweep",
"durationMs": 9690.432999999985,
"styleRecalcs": 249,
"styleRecalcDurationMs": 48.52400000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 4605.7300000000005,
"heapDeltaBytes": -3186824,
"heapUsedBytes": 75781336,
"domNodes": -251,
"jsHeapTotalBytes": 1495040,
"scriptDurationMs": 2075.755,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "viewport-pan-sweep",
"durationMs": 8374.054999999998,
"styleRecalcs": 250,
"styleRecalcDurationMs": 43.108,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 4123.8150000000005,
"heapDeltaBytes": -17836404,
"heapUsedBytes": 68446240,
"domNodes": -215,
"jsHeapTotalBytes": 5271552,
"scriptDurationMs": 1949.5059999999999,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "vue-large-graph-idle",
"durationMs": 12342.04699999998,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 11946.051000000001,
"heapDeltaBytes": -7306852,
"heapUsedBytes": 202310652,
"domNodes": -8061,
"jsHeapTotalBytes": -9183232,
"scriptDurationMs": 178.11900000000003,
"eventListeners": -14928,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.779999999999927,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-idle",
"durationMs": 12475.450000000023,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 12187.699,
"heapDeltaBytes": -41024744,
"heapUsedBytes": 191044756,
"domNodes": -8061,
"jsHeapTotalBytes": -11067392,
"scriptDurationMs": 172.51,
"eventListeners": -14928,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.223333333333358,
"p95FrameDurationMs": 16.80000000000291
},
{
"name": "vue-large-graph-pan",
"durationMs": 14848.204000000009,
"styleRecalcs": 72,
"styleRecalcDurationMs": 16.537999999999997,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 14735.413,
"heapDeltaBytes": -41122096,
"heapUsedBytes": 190964424,
"domNodes": -8061,
"jsHeapTotalBytes": -11661312,
"scriptDurationMs": 653.631,
"eventListeners": -14928,
"totalBlockingTimeMs": 47,
"frameDurationMs": 17.223333333333237,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-pan",
"durationMs": 14497.741999999904,
"styleRecalcs": 71,
"styleRecalcDurationMs": 14.816999999999997,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 14292.587999999998,
"heapDeltaBytes": -51768696,
"heapUsedBytes": 179665728,
"domNodes": -8061,
"jsHeapTotalBytes": -15785984,
"scriptDurationMs": 642.307,
"eventListeners": -14928,
"totalBlockingTimeMs": 1,
"frameDurationMs": 17.219999999999953,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-node-resize-workload",
"durationMs": 4443.774000000019,
"styleRecalcs": 20,
"styleRecalcDurationMs": 55.836999999999996,
"layouts": 20,
"layoutDurationMs": 23.428000000000004,
"taskDurationMs": 4393.549,
"heapDeltaBytes": -35497648,
"heapUsedBytes": 176449164,
"domNodes": -8062,
"jsHeapTotalBytes": 19193856,
"scriptDurationMs": 75.901,
"eventListeners": -14928,
"totalBlockingTimeMs": 91,
"frameDurationMs": 17.220000000000013,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "vue-node-resize-workload",
"durationMs": 4351.602999999955,
"styleRecalcs": 20,
"styleRecalcDurationMs": 64.74,
"layouts": 20,
"layoutDurationMs": 28.78499999999999,
"taskDurationMs": 4303.656,
"heapDeltaBytes": -35972336,
"heapUsedBytes": 189701900,
"domNodes": -8062,
"jsHeapTotalBytes": 10952704,
"scriptDurationMs": 72.326,
"eventListeners": -14928,
"totalBlockingTimeMs": 22,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "vue-zoom-out-idle",
"durationMs": 6260.343000000034,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 6172.658,
"heapDeltaBytes": -2115244,
"heapUsedBytes": 179966920,
"domNodes": 0,
"jsHeapTotalBytes": -15728640,
"scriptDurationMs": 89.509,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.220000000000073,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-zoom-out-idle",
"durationMs": 5977.708000000007,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 5582.027999999999,
"heapDeltaBytes": -9970464,
"heapUsedBytes": 183292948,
"domNodes": 0,
"jsHeapTotalBytes": -7340032,
"scriptDurationMs": 76.312,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.220000000000073,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "workflow-execution",
"durationMs": 439.77000000001,
"styleRecalcs": 16,
"styleRecalcDurationMs": 24.55,
"layouts": 5,
"layoutDurationMs": 2.34,
"taskDurationMs": 113.062,
"heapDeltaBytes": 5155348,
"heapUsedBytes": 70603528,
"domNodes": 151,
"jsHeapTotalBytes": 262144,
"scriptDurationMs": 11.449999999999998,
"eventListeners": 81,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.663333333333338,
"p95FrameDurationMs": 16.799999999999727
},
{
"name": "workflow-execution",
"durationMs": 72.27899999998044,
"styleRecalcs": 6,
"styleRecalcDurationMs": 11.477999999999998,
"layouts": 4,
"layoutDurationMs": 1.6769999999999998,
"taskDurationMs": 56.50199999999998,
"heapDeltaBytes": 3174264,
"heapUsedBytes": 68375916,
"domNodes": 103,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 6.269,
"eventListeners": 33,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000273
}
]
}🔬 E2E Coverage
Top 10 files by uncovered lines
|
…tuck (PM-1496) Adds a failing regression test (no fix) for part of the RCA on PM-1496: disconnecting an autogrow image slot (e.g. OpenAIGPTImageNodeV2's model.images.image_2) immediately after it was connected - before a render frame elapses - is misread by withComfyAutogrow's swap-detection guard in dynamicWidgets.ts as the tail of that same connect, so the scheduled spare-slot compaction (autogrowInputDisconnected) never runs. The group is left with a stale, disconnected slot the UI's growth logic doesn't reclaim, matching the "can't reconnect until a full refresh" symptom reported against this node's autogrow image inputs.
afa49eb to
8d43f92
Compare
withComfyAutogrow's swap-detection guard used a single rAF-bounded pendingConnection slot to decide whether a disconnect was the synchronous tail of a same-slot link replacement (a "swap") or a genuine, separate disconnect. Any disconnect landing on that slot within the same animation frame was treated as a swap tail and autogrowInputDisconnected was skipped - including when the slot's own connect had already fired and a user (or an agent applying CRDT-driven ops with no render yield between them) disconnected it moments later, before the frame elapsed. connectSlots (LGraphNode.ts) fires the old link's disconnect event strictly before the new link's connect event when replacing a link on the same slot, both synchronously. Track whether the connect event for the current pendingConnection slot has already fired (pendingConnectionSeen); only a disconnect arriving before that point is the tail of a swap. A disconnect arriving after is a distinct operation and now correctly triggers the scheduled spare-slot compaction.
There was a problem hiding this comment.
🔍 Cursor Review — Consolidated panel
Triggered by @christian-byrne.
Found 6 finding(s).
| Severity | Count |
|---|---|
| 🟠 High | 1 |
| 🟡 Medium | 3 |
| ⚪ Nit | 2 |
Panel: 8/8 reviewers contributed findings.
Codecov Report✅ All modified and coverable lines are covered by tests. @@ Coverage Diff @@
## main #18289 +/- ##
==========================================
+ Coverage 83.28% 84.27% +0.98%
==========================================
Files 2981 2280 -701
Lines 146683 129501 -17182
Branches 39783 34969 -4814
==========================================
- Hits 122161 109133 -13028
+ Misses 23241 20074 -3167
+ Partials 1281 294 -987
Flags with carried forward coverage won't be shown. Click here to find out more.
... and 709 files with indirect coverage changes 🚀 New features to boost your workflow:
|
The prior fix's swappingConnection flag was node-wide, not per-slot: while a same-slot swap on one input was resolving, a genuine connect landing on a *different* input of the same node hit the same guard and got silently dropped, so autogrow's growth bookkeeping never ran for it. Replace the boolean with swappingSlot (the specific slot the swap is for) and clear it deterministically when that slot's own matching connect event fires, instead of relying solely on a requestAnimationFrame boundary that can stay suspended indefinitely in a hidden background tab. Also, per Cursor review on this PR: - Add connectivity assertions (isInputConnected, getInputLink target_slot) to the disconnect-then-compact regression test so a fix that reaches the right slot count by dropping or mis-transplanting the surviving link would no longer pass. - Add a regression test pinning the cross-slot drop scenario above. - Rename the existing regression test: its title described the buggy (pre-fix) outcome while its assertion already checked the corrected one. - Fix two comments: one overstated pendingConnection's blast radius as module-private state instead of per-node closure state, the other mis-cited a line of prose from autogrowInputPersistence.spec.ts as that spec's test title. pnpm vitest run src/core/graph/widgets/dynamicWidgets.test.ts: 25 passed. eslint and vue-tsc --noEmit clean on both changed files.
christian-byrne
left a comment
There was a problem hiding this comment.
Immediate connect/disconnect/reconnect can remove the replacement link after the next frame. The retained blocker is reproduced on this head and passes on the pinned base. Unit coverage does not establish the separately owned original agent-reporter acceptance.
|
Merged after current-main qualification. Cloud inclusion and original agent acceptance remain separate. Full context for agent readersSeptember 22, 05:56:52 UTC: the reconnect repair landed as the verified merge. The exact-head approval is published here; all fourteen threads are resolved and no human correctness objection was dismissed. The approval was agent-authored using Christian's account, not an independent human peer review. The PR body records that distinction and the authorized moving-main judgment. Three individually verified browser cases passed once on a served tree identical to the reviewed head. Current main advanced with an interacting named-input repair; a fresh unpublished combination passed 242 ordinary tests plus one expected failure across six suites, using existing dependencies without installation. The head had 87 successful and 16 skipped checks/statuses. The staleness guard still refused; its refusal was retained rather than represented as passing. The final qualification is recorded in the PR body. The release mapper and coordinator received the merge receipt. Preserve their existing Cloud integration route. The excluded unpack case and original in-app agent journey remain unproved. No deployment, feature-flag change, or new source correction is requested here. Earlier dated requests below are historical and do not reopen completed fixes. Historical correction requests, supersededMain refresh received; 131 targeted tests pass. The assertion-style suggestion is fixed. Landing qualification remains pending; no new source correction requested. Full context for agent readersSeptember 22, 04:25 UTC: the writer merged main into the current head and replaced both conditional throw guards with An independent bounded source-only check is examining the intervening graph collaborators. The merge-readiness guard exited 3 because GitHub's shared GraphQL rate limit was exhausted before it could evaluate the gate. That is unavailable qualification, not a failed product test and not permission to bypass the guard. Current CI was pending at the last successful query. Existing review history and source/release ownership remain; no approval, objection dismissal, queue action, merge or replacement runtime follows from these checks. September 22, 03:52 UTC: The current-main guard refuses staleness: 36 commits behind main at qualification, zero direct file overlap, but graph/fixture collaborators changed. No moving-main exception is claimed. At the last check, 70 checks succeeded, 17 were skipped and five were nonterminal. Preserve the existing writer, review history and source lease. The next action is current-main compatibility and terminal checks, not repeating the fixed findings or starting another writer. No merge or runtime is requested by this update. The dated requests below are historical. Updated September 22, 03:04 UTC: The remaining suggestion asks for an occupied-input replacement through Please return the new exact head after normal-hook metadata correction. Do not bypass hooks, change human attribution, dismiss review objections or merge. If the correction requires unavailable dependencies or authority, report that prerequisite instead. Current-main qualification, terminal successful checks and original reporter acceptance remain separate. The original request below is retained as historical reproduction evidence, not a request to redo the fixed work. The blocking finding is independently reproduced on the reviewed source, with the identical sequence passing on the pinned base: use the real connection API to connect, disconnect and reconnect before any animation frame, then settle queued callbacks. The replacement disappears from the target input and graph lookup, and its source output is disconnected. A separate immediate second-input reconnect with the first input surviving confirms the same introduced failure. Please reconcile or invalidate stale disconnect work when a slot reconnects and retain regression coverage for replacement link identity, graph lookup and both endpoints. Keep a passing control that reconnects after compaction. The settled-first-connect variant already fails at the base and is a separate pre-existing issue, not evidence that this PR introduced every timing failure. The existing full unit file passed 25 tests but missed this sequence. Five-case timing coverage produced three failures/two passes on head versus two failures/three passes with only production source restored to base. The original added tests fail on base as intended, so preserve their regression protection too. Run with The two test-quality notes are compatible: add full real-API replacement coverage; optional intermediate-state simulation may remain if it uses the authoritative five-argument callback type. The other notes concern unsupported materialization claims and redundant narration, not an expansion into a new materialization implementation. Existing overlapping graph changes and their owners remain separate. Original reporter and release-target browser acceptance are not established by these unit tests. This is a correction request to the existing writer, not a second writer, assignment change, merge grant or deployment request. All six published inline bodies, anchors and review identity passed exact API readback and post-submit evidence validation. Glossary: compaction removes unused dynamic input slots; an animation frame is the deferred callback boundary used here to schedule that work. |
…M-1496) A connect/disconnect/reconnect sequence on the same autogrow slot, all in immediate succession, dropped the reconnect's link once the deferred disconnect compaction (autogrowInputDisconnected) ran: it unconditionally treated the slot as still empty and bubbled a later slot's link into it, discarding the fresh connection. Skip the compaction when the slot already holds a link by the time the deferred callback runs. Also addresses review nits: real 5-arg onConnectInput signature in the same-slot swap test instead of a one-arg unknown cast, trimmed narration comments in the test file, and removed a redundant comment on the swappingSlot branch in withComfyAutogrow. Claude-Session: https://claude.ai/code/session_01CoKpwnbrjRvnAyVd6dnPiq
|
Addressed your changes-requested review in 09dd915:
Full suite ( |
|
Requested by Christian Byrne · Slack thread CI failure:
|
…loss (PM-1496) The per-slot swap guard added in this PR for the disconnect/reconnect race also fixes a second known bug: LGraph.convertToSubgraph reconnects grouped output links with a synchronous connectSlots call per slot, and the old node-wide swappingConnection guard swallowed the genuine connect events for slots 2+ while slot 1's swap was in flight. The test.fail()-marked repro for that bug (added in #18179) now passes for real, so CI failed with "Expected to fail, but passed." Removing the stale annotation turns the repro into a real regression test. Claude-Session: https://claude.ai/code/session_01CoKpwnbrjRvnAyVd6dnPiq
Playwright videos for new testsDownload the recorded videos and HTML report Recorded projects: chromium Recorded test titles:
|
christian-byrne
left a comment
There was a problem hiding this comment.
The stale-compaction blocker and four ancillary findings are addressed. Source checks pass, and the selected browser contract passed on the CI merge with head-identical production/spec blobs. One nonblocking occupied-slot integration-coverage suggestion remains; original reporter acceptance, current-base qualification, CI failures and existing human review disposition are separate.
…t (PM-1496) Christian Byrne's review (comment 4068130853) noted the existing same-slot swap test only calls onConnectInput/onConnectionsChange directly to arm the intermediate swap state - it never installs the replacement link or delivers the matching connect callback through the real connectSlots path, so the occupied-slot replacement lifecycle stayed unverified. Add a real-API test that replaces an already-connected autogrow input via node.connect() (the actual connectSlots production path) while a genuine connect on a different slot lands in the exact window between the swap's disconnect and its matching connect event. Confirmed this fails (drops the second slot's growth) when swappingSlot is reverted to the prior node-wide boolean, and passes with the per-slot guard restored.
725bbb0 to
b8fb0f8
Compare
christian-byrne
left a comment
There was a problem hiding this comment.
The occupied-slot coverage request is addressed; all six prior findings are corrected.
Verification and remaining boundaries
The new test exercises replacement of an occupied input through the real connection API while a different slot connects. Direct installed unit tooling passed 27/27 tests. Changing the slot-specific guard back to a node-wide guard makes this exact case fail; restoration passes 27/27. Scoped lint and format checks pass. One nonblocking test-style finding follows.
Production and browser-spec files are unchanged from the previous reviewed revision. Earlier hosted browser evidence remains qualified evidence from that combined revision, not current-main execution or original reporter acceptance. This COMMENT neither dismisses existing objections nor qualifies landing; current checks and base compatibility require separate verification.
Glossary: combined revision means the PR merged with the main revision tested by CI.
…rrowing Replaces conditional-throw narrowing of oldLink/newLink in the occupied-slot swap regression test with Vitest's assert.exists, per the test readability contract.
…ect-pending-connection-race
christian-byrne
left a comment
There was a problem hiding this comment.
The six original findings and both follow-up requests are fixed at this head, including occupied-slot replacement coverage and assertion narrowing. All eight corresponding review threads are resolved after checking the source and author replies.
The verified browser shard individually passes plain conversion, autogrow loading and autogrow conversion on a generated merge whose full tree equals this head. Historical unit and mutation-test results retain their original source identities.
This approval is bound to that head and the reviewed base. It does not establish the excluded unpack case, original in-app agent acceptance, or qualification against newer main. CI means hosted automated checks; a generated merge combines the reviewed branch with its base.
|
|
… refresh (PM-1496) [cloud/1.54] (#18386) Reconnect repair qualified. Required checks passed. Authenticated reload remains open. <details><summary>Full context for agent readers</summary> Reviewed [head 1f35e0f](1f35e0f) against [tested Cloud target 9199835](9199835). The unmodified target-aware guard passed at that target: 76 successes, 12 skips, no unresolved threads or later objections. Fully paginated review and inline-comment collections are empty; issue comments contain only bots. No standing human approval is claimed. The [hosted unit job](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35700635510/job/106658136560) tested [merge b13fbb6](b13fbb6), whose parents are exactly that target and head. Its full tree equals the published head, 2ce4434. Results: 1552 frontend files, 21396 passes, six expected failures and seven skips. This includes dynamicWidgets 27, graphMutations 58, nodeDataStore 32 and materializer input-order two tests in the combined run. Required [lint](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35700635579/job/106657540942) and [browser aggregate](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35700635809/job/106664200450) passed. The newer [risk grading](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35700636057/job/106658217608) succeeded at this head; earlier cancelled risk jobs remain historical, not green evidence. The separate risk-publish step is skipped. The runtime change has the same stable patch identity as [merged source repair](#18289). The backport retained both release Pinia setup imports and the new assertion/callback imports. It preserves the source's immediate disconnect/reconnect and occupied-slot behavior, without importing unrelated main-branch changes. Prior browser evidence covers plain conversion, autogrow load and selected autogrow conversion only; unpack and authenticated in-app acceptance are not credited. The target then advanced by exactly [workflow carrier 18330](86d76db). The normal guard now refuses staleness; it has not passed at that newer target. There is zero file overlap: the target delta changes workflowStore.ts, its tests and test-only composer caret setup; this carrier changes dynamicWidgets.ts, its tests and the subgraph browser specification. The workflow delta filters missing tab entries and aligns tab reorder/navigation with that filtered list. It does not alter graph link storage, slot ordering, connection callbacks or this repair's frame lifecycle. The reconnect repair does not read those workflow-list APIs. A conflict-free merge-tree calculation produces 9757b73. This combined tree has not been executed as a new runtime; compatibility is the explicit inspected-delta judgment authorized by the coordinator, not a fabricated passing guard or repeated CI run. Justification: Necessary Cloud reconnect backport under Christian's explicit admin-merge grant and documented moving-target compatibility decision. Substantive source/backport review and exact-head required checks passed, with no recorded human objection. No new approval round or human approval is invented. Fresh target, head, reviews/comments and checks precede the merge. Authenticated edit/save/reload remains a separate open product blocker; landing does not authorize deployment or exposure widening. Carrier means release backport PR; patch identity compares code changes independently of line offsets; merge-tree checks textual integration, not runtime behavior. </details> <!-- authored-by:agent slack-claude-pilot --> <details><summary>Original author description (historical)</summary> <!-- ccr-slack-attribution --> _Requested by **Christian Byrne** · [Slack thread](https://comfy-organization.slack.com/archives/C0BGH348Z0C/p1789991789275729?thread_ts=1789991789.275729&cid=C0BGH348Z0C)_ Backport of [source reconnect repair](#18289) to `cloud/1.54`. Manual cherry-pick — the automated backport (run [35692714741](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35692714741)) reported a merge conflict in `src/core/graph/widgets/dynamicWidgets.test.ts`. ## Summary (from [source reconnect repair](#18289)) An autogrow image slot (e.g. `model.images.image_2`) that gets disconnected immediately after being connected — with no render frame in between, as happens when the in-app agent applies connect/disconnect CRDT ops back-to-back — was misclassified by `withComfyAutogrow`'s same-slot swap guard as the tail end of an atomic same-slot swap, so the real compaction was never scheduled. The slot stayed stuck, disconnected, in the middle of the group, and the UI refused a new link into it until a full page reload. The swap guard now tracks two things instead of relying only on "did a render frame elapse": whether the slot's own connect event has already fired (`pendingConnectionSeen`), and which specific slot is mid-swap (`swappingSlot`, scoped per-slot rather than node-wide). A disconnect arriving after its slot's connect event already fired is now correctly treated as a separate, later operation, so `autogrowInputDisconnected` runs and the group compacts back down. Scoping the guard per-slot also stops a genuine connect on a *different* slot of the same node from being mistaken for the swap's own connect and silently dropped. ## Conflict resolution Only `src/core/graph/widgets/dynamicWidgets.test.ts` conflicted, in the import block: - `cloud/1.54`'s copy of this test file already sets up `createTestingPinia`/`setActivePinia` (a `beforeEach` Pinia reset used by describe blocks not present in [source reconnect repair](#18289 diff), which `main`'s copy at the time of [source reconnect repair](#18289) didn't need. - [source reconnect repair](#18289) added `assert` (used at `assert.exists(oldLink)` / `assert.exists(newLink)`) and `useChainCallback` (used to compose `node.onConnectionsChange` in the new tests) to the `vitest`/local imports. Resolved by keeping both import sets — the existing Pinia setup imports plus the new `assert`/`useChainCallback` imports — with no other changes. `src/core/graph/widgets/dynamicWidgets.ts` (the runtime fix) and `browser_tests/tests/subgraph/subgraphConvertAutogrowInputs.spec.ts` auto-merged cleanly with no conflicts; nothing else from `main` was pulled in. No lockfile or `package.json` changes were needed. ## Verification - `pnpm vitest run src/core/graph/widgets/dynamicWidgets.test.ts` — 27 passed. - `pnpm typecheck` and `pnpm typecheck:browser` — passed (via pre-commit hook). - `pnpm exec eslint --fix` / `pnpm exec oxlint --type-aware --fix` / `pnpm exec oxfmt --write` on changed files — clean, no warnings. - `.github/scripts/check-ai-co-authors.sh` against the merge-base with `origin/cloud/1.54` — no AI co-author trailers (the inherited `Co-authored-by: Claude <noreply@anthropic.com>` trailer from the `main` squash commit was stripped via `git commit --amend`). ## Coverage note The browser test spec (`browser_tests/tests/subgraph/subgraphConvertAutogrowInputs.spec.ts`) carried over cleanly with its full diff — no browser coverage was dropped in this backport. Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01CoKpwnbrjRvnAyVd6dnPiq --- _Generated by [Claude Code](https://claude.ai/code/session_01CoKpwnbrjRvnAyVd6dnPiq)_ </details> Co-authored-by: claude[bot] <209825114+claude[bot]@users.noreply.github.com> Co-authored-by: Claude <noreply@anthropic.com>
… refresh (PM-1496) [cloud/1.54] (#18386) Reconnect repair qualified. Required checks passed. Authenticated reload remains open. <details><summary>Full context for agent readers</summary> Reviewed [head 1f35e0f](1f35e0f) against [tested Cloud target 9199835](9199835). The unmodified target-aware guard passed at that target: 76 successes, 12 skips, no unresolved threads or later objections. Fully paginated review and inline-comment collections are empty; issue comments contain only bots. No standing human approval is claimed. The [hosted unit job](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35700635510/job/106658136560) tested [merge b13fbb6](b13fbb6), whose parents are exactly that target and head. Its full tree equals the published head, 2ce4434. Results: 1552 frontend files, 21396 passes, six expected failures and seven skips. This includes dynamicWidgets 27, graphMutations 58, nodeDataStore 32 and materializer input-order two tests in the combined run. Required [lint](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35700635579/job/106657540942) and [browser aggregate](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35700635809/job/106664200450) passed. The newer [risk grading](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35700636057/job/106658217608) succeeded at this head; earlier cancelled risk jobs remain historical, not green evidence. The separate risk-publish step is skipped. The runtime change has the same stable patch identity as [merged source repair](#18289). The backport retained both release Pinia setup imports and the new assertion/callback imports. It preserves the source's immediate disconnect/reconnect and occupied-slot behavior, without importing unrelated main-branch changes. Prior browser evidence covers plain conversion, autogrow load and selected autogrow conversion only; unpack and authenticated in-app acceptance are not credited. The target then advanced by exactly [workflow carrier 18330](86d76db). The normal guard now refuses staleness; it has not passed at that newer target. There is zero file overlap: the target delta changes workflowStore.ts, its tests and test-only composer caret setup; this carrier changes dynamicWidgets.ts, its tests and the subgraph browser specification. The workflow delta filters missing tab entries and aligns tab reorder/navigation with that filtered list. It does not alter graph link storage, slot ordering, connection callbacks or this repair's frame lifecycle. The reconnect repair does not read those workflow-list APIs. A conflict-free merge-tree calculation produces 9757b73. This combined tree has not been executed as a new runtime; compatibility is the explicit inspected-delta judgment authorized by the coordinator, not a fabricated passing guard or repeated CI run. Justification: Necessary Cloud reconnect backport under Christian's explicit admin-merge grant and documented moving-target compatibility decision. Substantive source/backport review and exact-head required checks passed, with no recorded human objection. No new approval round or human approval is invented. Fresh target, head, reviews/comments and checks precede the merge. Authenticated edit/save/reload remains a separate open product blocker; landing does not authorize deployment or exposure widening. Carrier means release backport PR; patch identity compares code changes independently of line offsets; merge-tree checks textual integration, not runtime behavior. </details> <!-- authored-by:agent slack-claude-pilot --> <details><summary>Original author description (historical)</summary> <!-- ccr-slack-attribution --> _Requested by **Christian Byrne** · [Slack thread](https://comfy-organization.slack.com/archives/C0BGH348Z0C/p1789991789275729?thread_ts=1789991789.275729&cid=C0BGH348Z0C)_ Backport of [source reconnect repair](#18289) to `cloud/1.54`. Manual cherry-pick — the automated backport (run [35692714741](https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/35692714741)) reported a merge conflict in `src/core/graph/widgets/dynamicWidgets.test.ts`. ## Summary (from [source reconnect repair](#18289)) An autogrow image slot (e.g. `model.images.image_2`) that gets disconnected immediately after being connected — with no render frame in between, as happens when the in-app agent applies connect/disconnect CRDT ops back-to-back — was misclassified by `withComfyAutogrow`'s same-slot swap guard as the tail end of an atomic same-slot swap, so the real compaction was never scheduled. The slot stayed stuck, disconnected, in the middle of the group, and the UI refused a new link into it until a full page reload. The swap guard now tracks two things instead of relying only on "did a render frame elapse": whether the slot's own connect event has already fired (`pendingConnectionSeen`), and which specific slot is mid-swap (`swappingSlot`, scoped per-slot rather than node-wide). A disconnect arriving after its slot's connect event already fired is now correctly treated as a separate, later operation, so `autogrowInputDisconnected` runs and the group compacts back down. Scoping the guard per-slot also stops a genuine connect on a *different* slot of the same node from being mistaken for the swap's own connect and silently dropped. ## Conflict resolution Only `src/core/graph/widgets/dynamicWidgets.test.ts` conflicted, in the import block: - `cloud/1.54`'s copy of this test file already sets up `createTestingPinia`/`setActivePinia` (a `beforeEach` Pinia reset used by describe blocks not present in [source reconnect repair](#18289 diff), which `main`'s copy at the time of [source reconnect repair](#18289) didn't need. - [source reconnect repair](#18289) added `assert` (used at `assert.exists(oldLink)` / `assert.exists(newLink)`) and `useChainCallback` (used to compose `node.onConnectionsChange` in the new tests) to the `vitest`/local imports. Resolved by keeping both import sets — the existing Pinia setup imports plus the new `assert`/`useChainCallback` imports — with no other changes. `src/core/graph/widgets/dynamicWidgets.ts` (the runtime fix) and `browser_tests/tests/subgraph/subgraphConvertAutogrowInputs.spec.ts` auto-merged cleanly with no conflicts; nothing else from `main` was pulled in. No lockfile or `package.json` changes were needed. ## Verification - `pnpm vitest run src/core/graph/widgets/dynamicWidgets.test.ts` — 27 passed. - `pnpm typecheck` and `pnpm typecheck:browser` — passed (via pre-commit hook). - `pnpm exec eslint --fix` / `pnpm exec oxlint --type-aware --fix` / `pnpm exec oxfmt --write` on changed files — clean, no warnings. - `.github/scripts/check-ai-co-authors.sh` against the merge-base with `origin/cloud/1.54` — no AI co-author trailers (the inherited `Co-authored-by: Claude <noreply@anthropic.com>` trailer from the `main` squash commit was stripped via `git commit --amend`). ## Coverage note The browser test spec (`browser_tests/tests/subgraph/subgraphConvertAutogrowInputs.spec.ts`) carried over cleanly with its full diff — no browser coverage was dropped in this backport. Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01CoKpwnbrjRvnAyVd6dnPiq --- _Generated by [Claude Code](https://claude.ai/code/session_01CoKpwnbrjRvnAyVd6dnPiq)_ </details> (cherry picked from commit b06429b)
## Summary Ports to `core/1.54` every merged `main` Agent fix that `cloud/1.54` carries and core does not (the Agent UI catch-up from #18622, the cloud-only CRDT and workflow fixes, and the round-2 fixes), so both release lines are symmetric ahead of the local-agent rollout. ## Changes - **What**: `cherry-pick -x`, in three groups: - Agent UI, as in #18622: #17339, #17908, #17420, #18348, #18043, #17621 (without its two `src/lib/litegraph/` files; a 5 px HUD offset in `SideToolbar.vue` compensates), #17471, #17690. - Cloud-only fixes: #18063, #18073, #18064, #18033, #18289, #18604, #18616, #18376. - Round 2: #17568, #17628, #16491, #18453, #17767, #18431 (from #18571), #18603. - **Already on `core/1.54`**: #17439 (#18161), #17912 (#18186), #18391 (#18595). - **Excluded**: #17730 (core lacks its three files), #18421 (dead code on core builds), #16917 (conflicts with later core changes), #18066 (Christian's request), #18375 and #18251 (need their litegraph hunks). ## Review Focus - On this line the Agent panel registers only in cloud builds (`src/extensions/core/index.ts:41-44`); the shared chrome from #17621 and #17471 and the host telemetry from #18376 and #18603 ship in every build. - `HostTelemetrySink.ts` and its test equal `main`'s files (#18376, then #18603 on top). - #18064: `agentConversationHostDoc.test.ts` keeps this line's file; #18063's new test uses `main`'s version (multi-player 0.3.x stamps). - Supersedes #18571 and the closed #18637. Linear: [PM-1657](https://linear.app/comfyorg/issue/PM-1657), [PM-1656](https://linear.app/comfyorg/issue/PM-1656) ## Screenshots (if applicable) Visual changes are the ones reviewed on `main` and carried to cloud in #18622. --------- Co-authored-by: uytieu <uytieu@users.noreply.github.com> Co-authored-by: ryan-di <ryan.weihao.di@gmail.com> Co-authored-by: jaeone lee <89377375+jaeone94@users.noreply.github.com> Co-authored-by: Christian Byrne <cbyrne@comfy.org> Co-authored-by: Amp <amp@ampcode.com> Co-authored-by: github-actions <github-actions@github.com> Co-authored-by: Alexander Brown <drjkl@comfy.org> Co-authored-by: Comfy Org PR Bot <snomiao+comfy-pr@gmail.com> Co-authored-by: claude[bot] <209825114+claude[bot]@users.noreply.github.com> Co-authored-by: Hunter <huntcsg@users.noreply.github.com> Co-authored-by: Maanil Verma <vermaMaanil97@gmail.com>
|
|
|
|
…p to a subgraph (FE-3325) (Comfy-Org#20743) Fixes FE-3325 (Pylon#20157): Convert to Subgraph put links on the wrong inputs when a node's autogrow group had grown past its default size. - **Fix:** `multiClone` clones the live node's serialization directly instead of round-tripping it through a fresh, graph-less node. - **Proof:** the existing FE-3325 vitest and Playwright repros now run as plain tests. They fail on the parent with the original mis-slotting and pass with the fix. - **Scope:** one function in `subgraphUtils.ts`. `convertToSubgraph` is its only caller. <details><summary>Full context for agent readers</summary> ## Bug When a node's `COMFY_AUTOGROW_V3` group has grown past its default size, Convert to Subgraph puts the links of every input after the group onto the wrong slots. Example: a node with `images` (autogrow), then `mask`, then `strength`, with `image0`–`image2`, `mask` and `strength` connected. Inside the new subgraph, before this fix: | input | expected link | actual link | |---|---|---| | `images.image2` | IMAGE | **MASK** | | `mask` | MASK | **IMAGE** | | `images.image3` | – | **FLOAT** | | `strength` | FLOAT | **–** | ## Root cause 1. `LGraph._convertToSubgraphImpl` calls `multiClone(nodes)`. `multiClone` (`src/lib/litegraph/src/subgraph/subgraphUtils.ts`) did `createNode(type)` → `newNode.configure(data)` → `newNode.serialize()`. 2. The fresh node only has the default autogrow layout. `ComfyNode.configure` (`src/services/litegraphService.ts`) treats the grown slots as "extra" inputs and appends them after `mask`/`strength`. 3. The fresh node has no graph, so `input.link`, which is now read from the link store, serializes as `null`. Autogrow cannot regrow from the replayed connections. 4. `createSubgraph` builds the links using the original `target_slot`s. With no input link ids, `realignInputLinkSlots` has nothing to match by name, so the stale indices are kept. The round trip predates the link store. When `input.link` was a plain field, it survived configure → serialize, so the round trip used to keep link ids. It no longer does. ## Fix `multiClone` now returns `structuredClone(node.serialize())` for each node. The serialized inputs keep their live order and link ids, so the existing name-based realignment in `LGraph.configure` repoints every link. The `createNode` failure branch is gone because the direct clone already covers it; that was the only use of the `LiteGraph` import. ## Tests - `src/core/graph/widgets/autogrowConvertToSubgraph.test.ts`: the two `test.fails` link-mapping tests (boundary links and internal links) are now plain `test`s. The `currently mis-slots` tests and `MIS_SLOTTED_CONNECTIONS` are deleted because they pinned the bug's wrong mapping. No assertion of the correct mapping was removed. - `browser_tests/tests/subgraph/subgraphConvertAutogrowSlotOrder.spec.ts`: `test.fail()` removed. The spec loads `subgraphs/autogrow-then-mask`, wires a grown group plus `mask`, converts through the selection toolbox, and asserts every inner link type by input name. Its helpers live in `browser_tests/fixtures/utils/autogrowConvertToSubgraph.ts`. - Review follow-up (`1cfc3422bc`): the unit test now feeds each IMAGE input from its own source output and asserts the origin output per input, so a swap between IMAGE links fails too. Boundary links are traced through the subgraph host's input. Still red on the parent (both link-mapping tests) and green on the fix; the Playwright spec is still red on the parent (`images.image3: MASK`, `mask: IMAGE`) and green on the fix. ## Verification Run locally on this branch (merged with `main` at `8bf8cf6f8e`): - **Red on the parent.** With only `subgraphUtils.ts` reverted: - vitest: both link-mapping tests fail with `images.image2: MASK`, `images.image3: FLOAT`, `mask: IMAGE`. - Playwright (chromium, local backend + dev server): the spec fails with `images.image3: MASK`, `mask: IMAGE`. - **Green on the fix.** - vitest: the FE-3325 file gives 5 passed, and `src/lib/litegraph/src/subgraph` gives 375 passed in total. - Full unit suite: 31,541 passed. The only failures were 2 to 3 git-driven tests in `scripts/server-fact-gate.test.ts`, which time out at 5 s on this host whether run alone or with the suite. They do not import the changed code. - Playwright: all 20 spec files that convert to a subgraph give 142 passed. That covers `subgraphConvertAutogrowInputs`, `subgraphCreationBoundaryLinks`, `subgraphPromotion`, `subgraphNestedConversionValuePreservation`, `dynamicGroup`, `nativeLoraStack`, `changeTracker` and others. Two promotion tests timed out in fixture setup before their bodies ran on the first pass, and both passed on rerun. - `pnpm typecheck` and `pnpm typecheck:browser` pass, as do oxfmt and oxlint. The pre-push hooks (depcruise, knip, fallow audit) passed on the push. ## Related work - **Comfy-Org#21298** (draft, FE-3463, DynamicCombo): touches the same `multiClone` round trip and keeps it, restoring link ids by name afterwards. Its 25-test `dynamicInputsConvertToSubgraph.test.ts` also passes against this branch's `multiClone`, and this PR's tests pass with Comfy-Org#21298's fix. Proposed resolution, posted on Comfy-Org#21298: this PR's `multiClone` is the single fix, and Comfy-Org#21298 keeps only its tests. - **Comfy-Org#16897** (merged): only fixes *unpack*. - **Comfy-Org#15414** (merged): repro for dropped boundary links, fixed by PM-1496 Comfy-Org#18289. Its fixture has no linked input after the grown group. - **Comfy-Org#20234** (open, FE-2443): fixes the *workflow reload* path. Its realign returns early when `!node.graph`, so it never reaches `multiClone`. - **Comfy-Org#19625** (open): prunes dangling inner links. Doesn't touch `multiClone`. ## History This PR started as a repro-only handoff (expected-failure tests, no fix), at the request of the FE-3325 reporter. The fix was then added on top, so the red proof and the fix land as one change instead of the expected-failure tests merging as if they resolved the bug. Screenshots are not included because the bug is in link-to-slot data, which the tests read directly from the graph. </details> --------- Co-authored-by: Glary-Bot <glary-bot@users.noreply.github.com>
<!-- ccr-slack-attribution -->
Requested by Denys Puziak · Slack thread
Before
Denys wired an IMAGE output straight into
OpenAIGPTImageNodeV2's STRINGpromptinput and asked the in-app agent to fix it (reconnect report PM-1496 and parent agent-wiring report PM-1145). While investigating, we found that after an autogrow image slot (e.g.model.images.image_2) gets disconnected immediately after being connected - with no render frame in between, as happens when the in-app agent applies connect/disconnect CRDT ops back-to-back -withComfyAutogrow's same-slot swap guard insrc/core/graph/widgets/dynamicWidgets.tsmisclassified the disconnect as the tail end of an atomic same-slot swap and skipped scheduling the real compaction. The slot was left stuck, disconnected, in the middle of the group, and the UI silently refused to accept a new link into it until the page was fully reloaded.After
The swap guard now tracks two things instead of relying only on "did a render frame elapse": whether the slot's own connect event has already fired (
pendingConnectionSeen), and which specific slot is mid-swap (swappingSlot, scoped per-slot rather than node-wide). A disconnect that arrives after its slot's connect event already fired is correctly treated as a separate, later operation rather than a swap, soautogrowInputDisconnectedruns and the group compacts back down to its one spare slot. Scoping the guard to a single slot also means a genuine connect on a different slot of the same node is never mistaken for a swap's own connect and silently dropped.Covered by two new tests in
src/core/graph/widgets/dynamicWidgets.test.ts:eslint/oxfmtclean on the changed files.Generated with Claude Code
https://claude.ai/code/session_01CoKpwnbrjRvnAyVd6dnPiq
Generated by Claude Code
Justification: Christian authorized Agent-release source completion and qualified admin merges on September 22. The current-head approval was produced by agents using Christian's account, not an independent human peer review. No outstanding human correctness objection was found. Three individually verified browser cases passed on a served tree identical to the reviewed head; unpack and original in-app agent acceptance remain unproven. Post-merge human peer review remains outstanding.
Full context for agent readers
Moving-main merge judgment: the reviewed head is fourteen commits behind main at qualification. Complete check pagination found 87 successful and 16 skipped check/status entries, with none adverse or pending. The merge guard refused staleness; it did not pass. The overlapping
dynamicWidgets.tschange captures the existing graph reference insidewithComfyMatchTypewithout changing the autogrow lifecycle guards. The intervening slot synchronization, widget/serialization and asset-drag changes were inspected. A fresh unpublished merge with that main passed 170 targeted tests in eight graph, materializer, slot and widget suites, using the lane's existing dependencies without installation. Earlier prospective-main checks passed another 239 and 44 tests. These checks support this bounded source landing; they are not hosted browser execution on the newer main or deployment acceptance.All fourteen review threads are resolved; the earlier changes-requested review was agent-authored and superseded by the exact-head approval. No human review was dismissed. The selected autogrow conversion, autogrow image loading and ordinary multi-input conversion each passed once without retry in the hosted shard. The custom-node unpack case was excluded, not passed. Cloud release inclusion and original user-journey acceptance remain separate obligations.
Final main refresh adds one website-only revert, making fifteen commits behind. Its complete path list is confined to
apps/website/and.github/pr-images/, with no application runtime, build configuration or dependency change. This disjoint delta does not invalidate the targeted prospective-merge tests. Final head/checks/reviews remain unchanged; the guard still refuses staleness only.Subsequent identity refresh stopped the merge before execution because named-input preservation landed, making sixteen commits behind. Unlike the website delta, this changes interacting agent materialization and remote slot synchronization. After inspecting both production changes, a new unpublished combination passed 242 ordinary tests and one expected failure across six suites: autogrow, node data, materialization, materializer input ordering, graph mutations and the agent follower. The new code preserves named live input slots and realigns saved links after configuration; it does not replace the synchronous connection callbacks whose swap-state tracking this PR repairs. This supersedes the earlier moving-main qualification, with the same existing-dependency and browser-evidence limitations.
Glossary: prospective merge means an unpublished combination tested locally; tree-identical means all tracked files match; source landing does not mean a production deployment.