Repository navigation
[chore] Update Ingest API types from cloud@44b2ac7 - #19524
Conversation
🌐 Website E2ETip All tests passed.
🎭 Playwright: ✅ 2146 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 a9968daf959f722c7461c3a48efee1d8b5404d91
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 Size
⚡ Performance Report
Absolute values
Raw data{
"timestamp": "2026-10-03T00:29:15.263Z",
"gitSha": "95a192bd8117276105fddf51ddb5be7eb5d17849",
"branch": "update-ingest-types-cloud-1.55",
"measurements": [
{
"name": "canvas-idle",
"durationMs": 1999.0179999999782,
"styleRecalcs": 12,
"styleRecalcDurationMs": 9.464,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 456.20699999999994,
"heapDeltaBytes": 7104000,
"heapUsedBytes": 75257944,
"domNodes": -256,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 8.733,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-idle",
"durationMs": 2006.042999999977,
"styleRecalcs": 10,
"styleRecalcDurationMs": 9.350000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 465.83699999999993,
"heapDeltaBytes": 2182420,
"heapUsedBytes": 69998640,
"domNodes": -262,
"jsHeapTotalBytes": 4972544,
"scriptDurationMs": 8.351,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1737.9250000000184,
"styleRecalcs": 45,
"styleRecalcDurationMs": 13.965000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 698.226,
"heapDeltaBytes": -6348416,
"heapUsedBytes": 61522420,
"domNodes": -259,
"jsHeapTotalBytes": 4448256,
"scriptDurationMs": 158.856,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1739.512999999988,
"styleRecalcs": 45,
"styleRecalcDurationMs": 13.780000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 803.728,
"heapDeltaBytes": -6225332,
"heapUsedBytes": 56457804,
"domNodes": -262,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 165.413,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1666.1530000000084,
"styleRecalcs": 32,
"styleRecalcDurationMs": 15.270000000000001,
"layouts": 6,
"layoutDurationMs": 1.029,
"taskDurationMs": 343.68,
"heapDeltaBytes": 3146836,
"heapUsedBytes": 70882472,
"domNodes": 80,
"jsHeapTotalBytes": 6029312,
"scriptDurationMs": 10.786999999999997,
"eventListeners": 15,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1665.091000000075,
"styleRecalcs": 32,
"styleRecalcDurationMs": 12.780999999999999,
"layouts": 6,
"layoutDurationMs": 1.0040000000000002,
"taskDurationMs": 325.032,
"heapDeltaBytes": 2912780,
"heapUsedBytes": 71189660,
"domNodes": 77,
"jsHeapTotalBytes": 4980736,
"scriptDurationMs": 10.225000000000001,
"eventListeners": 15,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "dom-widget-clipping",
"durationMs": 624.7589999999832,
"styleRecalcs": 14,
"styleRecalcDurationMs": 9.568000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 461.022,
"heapDeltaBytes": 18679364,
"heapUsedBytes": 86483064,
"domNodes": 24,
"jsHeapTotalBytes": 4718592,
"scriptDurationMs": 176.69199999999998,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-clipping",
"durationMs": 644.1730000000234,
"styleRecalcs": 12,
"styleRecalcDurationMs": 7.917999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 461.939,
"heapDeltaBytes": 18613436,
"heapUsedBytes": 86539564,
"domNodes": 20,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 175.19600000000003,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.670000000000012,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-order-dense-n200-w200-i120",
"durationMs": 1222.879999999975,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 913.8689999999999,
"heapDeltaBytes": -17071172,
"heapUsedBytes": 73244436,
"domNodes": -295,
"jsHeapTotalBytes": 20054016,
"scriptDurationMs": 75.174,
"eventListeners": -217,
"totalBlockingTimeMs": 567,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "dom-widget-order-dense-n200-w200-i120",
"durationMs": 1243.1379999999308,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 901.51,
"heapDeltaBytes": 14044208,
"heapUsedBytes": 104664668,
"domNodes": -296,
"jsHeapTotalBytes": 21909504,
"scriptDurationMs": 70.31599999999999,
"eventListeners": -195,
"totalBlockingTimeMs": 615,
"frameDurationMs": 16.66999999999998,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "dom-widget-order-sparse-n1000-w20-i120",
"durationMs": 842.0249999999783,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 457.68699999999995,
"heapDeltaBytes": 15724672,
"heapUsedBytes": 91248124,
"domNodes": 0,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 161.13299999999998,
"eventListeners": 0,
"totalBlockingTimeMs": 222,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-order-sparse-n1000-w20-i120",
"durationMs": 844.7629999999435,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 466.62800000000004,
"heapDeltaBytes": 13933464,
"heapUsedBytes": 90539292,
"domNodes": 0,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 165.669,
"eventListeners": 0,
"totalBlockingTimeMs": 229,
"frameDurationMs": 16.670000000000012,
"p95FrameDurationMs": 16.80000000000109
},
{
"name": "extension-clean-combined-draw",
"durationMs": 10623.270999999988,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 9618.849,
"heapDeltaBytes": 11288436,
"heapUsedBytes": 93530468,
"domNodes": -16,
"jsHeapTotalBytes": 6815744,
"scriptDurationMs": 8242.205,
"eventListeners": 0,
"totalBlockingTimeMs": 1465,
"frameDurationMs": 18.890000000000025,
"p95FrameDurationMs": 50
},
{
"name": "extension-clean-combined-draw",
"durationMs": 10691.845999999941,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 9686.238,
"heapDeltaBytes": -6962544,
"heapUsedBytes": 75393972,
"domNodes": -16,
"jsHeapTotalBytes": 9961472,
"scriptDurationMs": 8267.287,
"eventListeners": 0,
"totalBlockingTimeMs": 1544,
"frameDurationMs": 19.996666666666655,
"p95FrameDurationMs": 50
},
{
"name": "extension-rgthree-prototype-draw",
"durationMs": 10091.486999999972,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 9166.008,
"heapDeltaBytes": 183328,
"heapUsedBytes": 82397308,
"domNodes": -14,
"jsHeapTotalBytes": 6291456,
"scriptDurationMs": 7798.027,
"eventListeners": 0,
"totalBlockingTimeMs": 1111,
"frameDurationMs": 18.889999999999965,
"p95FrameDurationMs": 50
},
{
"name": "extension-rgthree-prototype-draw",
"durationMs": 10422.35500000004,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 9592.367999999999,
"heapDeltaBytes": -5327144,
"heapUsedBytes": 72057216,
"domNodes": -270,
"jsHeapTotalBytes": 1040384,
"scriptDurationMs": 7886.380999999999,
"eventListeners": -174,
"totalBlockingTimeMs": 1043,
"frameDurationMs": 20,
"p95FrameDurationMs": 49.899999999999636
},
{
"name": "large-graph-idle",
"durationMs": 2029.9679999999967,
"styleRecalcs": 9,
"styleRecalcDurationMs": 7.226,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 598.261,
"heapDeltaBytes": -3888548,
"heapUsedBytes": 83886884,
"domNodes": -240,
"jsHeapTotalBytes": -495616,
"scriptDurationMs": 33.824999999999996,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-idle",
"durationMs": 2040.4310000000123,
"styleRecalcs": 8,
"styleRecalcDurationMs": 7.338999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 588.8200000000002,
"heapDeltaBytes": -3561400,
"heapUsedBytes": 83866644,
"domNodes": -240,
"jsHeapTotalBytes": -1019904,
"scriptDurationMs": 32.355,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "large-graph-pan",
"durationMs": 2347.8039999999964,
"styleRecalcs": 69,
"styleRecalcDurationMs": 13.882000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1694.136,
"heapDeltaBytes": -13153316,
"heapUsedBytes": 67493388,
"domNodes": -256,
"jsHeapTotalBytes": -8192,
"scriptDurationMs": 905.906,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-pan",
"durationMs": 2309.53599999998,
"styleRecalcs": 68,
"styleRecalcDurationMs": 13.329000000000004,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1647.203,
"heapDeltaBytes": 15735428,
"heapUsedBytes": 80275928,
"domNodes": -246,
"jsHeapTotalBytes": 1273856,
"scriptDurationMs": 855.962,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-zoom",
"durationMs": 3453.950999999961,
"styleRecalcs": 62,
"styleRecalcDurationMs": 10.503000000000002,
"layouts": 60,
"layoutDurationMs": 12.596,
"taskDurationMs": 1655.5539999999999,
"heapDeltaBytes": 8333584,
"heapUsedBytes": 74452644,
"domNodes": 4,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 784.474,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-zoom",
"durationMs": 3536.4550000000463,
"styleRecalcs": 62,
"styleRecalcDurationMs": 10.825000000000001,
"layouts": 60,
"layoutDurationMs": 12.786999999999999,
"taskDurationMs": 1680.778,
"heapDeltaBytes": 7818488,
"heapUsedBytes": 74575720,
"domNodes": 4,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 805.191,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "legacy-node-drag",
"durationMs": 3165.4960000000187,
"styleRecalcs": 42,
"styleRecalcDurationMs": 6.699000000000003,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1924.0009999999997,
"heapDeltaBytes": 23251028,
"heapUsedBytes": 89676192,
"domNodes": 4,
"jsHeapTotalBytes": 942080,
"scriptDurationMs": 1010.0830000000001,
"eventListeners": 180,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.80000000000109
},
{
"name": "legacy-node-drag",
"durationMs": 3127.701000000002,
"styleRecalcs": 43,
"styleRecalcDurationMs": 6.677999999999996,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1982.989,
"heapDeltaBytes": 26695104,
"heapUsedBytes": 92463352,
"domNodes": -250,
"jsHeapTotalBytes": 679936,
"scriptDurationMs": 1014.357,
"eventListeners": 180,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666636,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "minimap-idle",
"durationMs": 2041.8280000000095,
"styleRecalcs": 8,
"styleRecalcDurationMs": 8.655,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 636.293,
"heapDeltaBytes": -8818512,
"heapUsedBytes": 80534672,
"domNodes": -250,
"jsHeapTotalBytes": -233472,
"scriptDurationMs": 40.60700000000002,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "minimap-idle",
"durationMs": 2001.1739999999918,
"styleRecalcs": 5,
"styleRecalcDurationMs": 3.733,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 504.815,
"heapDeltaBytes": 20092576,
"heapUsedBytes": 85000760,
"domNodes": 10,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 31.42199999999999,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 590.9469999999715,
"styleRecalcs": 47,
"styleRecalcDurationMs": 10.395999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 397.567,
"heapDeltaBytes": 13809276,
"heapUsedBytes": 83843124,
"domNodes": 20,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 146.761,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 579.3290000000297,
"styleRecalcs": 47,
"styleRecalcDurationMs": 11.61,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 410.522,
"heapDeltaBytes": 13983268,
"heapUsedBytes": 84238480,
"domNodes": 20,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 148.14300000000003,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-idle",
"durationMs": 2020.1159999999732,
"styleRecalcs": 11,
"styleRecalcDurationMs": 8.815000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 458.913,
"heapDeltaBytes": -2158792,
"heapUsedBytes": 68009696,
"domNodes": -275,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 8.187999999999997,
"eventListeners": -220,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "subgraph-idle",
"durationMs": 2024.0780000000314,
"styleRecalcs": 9,
"styleRecalcDurationMs": 9.260000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 527.311,
"heapDeltaBytes": 3765840,
"heapUsedBytes": 67887892,
"domNodes": -278,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 10.313000000000002,
"eventListeners": -220,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1727.0809999999983,
"styleRecalcs": 46,
"styleRecalcDurationMs": 15.027,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 722.7640000000001,
"heapDeltaBytes": 6207212,
"heapUsedBytes": 76701860,
"domNodes": -275,
"jsHeapTotalBytes": 6545408,
"scriptDurationMs": 101.586,
"eventListeners": -198,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1723.5379999999623,
"styleRecalcs": 45,
"styleRecalcDurationMs": 16.006,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 753.6550000000001,
"heapDeltaBytes": 8399364,
"heapUsedBytes": 78403824,
"domNodes": -276,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 108.089,
"eventListeners": -198,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-transition-enter",
"durationMs": 1384.8169999999982,
"styleRecalcs": 17,
"styleRecalcDurationMs": 28.653000000000006,
"layouts": 5,
"layoutDurationMs": 13.008,
"taskDurationMs": 913.0619999999998,
"heapDeltaBytes": -6292736,
"heapUsedBytes": 83483180,
"domNodes": 13667,
"jsHeapTotalBytes": 10747904,
"scriptDurationMs": 29.429999999999996,
"eventListeners": 2383,
"totalBlockingTimeMs": 140,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "viewport-pan-sweep",
"durationMs": 8460.974000000022,
"styleRecalcs": 251,
"styleRecalcDurationMs": 39.345000000000006,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 5790.35,
"heapDeltaBytes": -1998600,
"heapUsedBytes": 85156044,
"domNodes": -242,
"jsHeapTotalBytes": 253952,
"scriptDurationMs": 2817.149,
"eventListeners": -152,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.670000000000012,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "viewport-pan-sweep",
"durationMs": 8407.493000000046,
"styleRecalcs": 250,
"styleRecalcDurationMs": 34.746,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 5430.072000000001,
"heapDeltaBytes": 417652,
"heapUsedBytes": 82288820,
"domNodes": -239,
"jsHeapTotalBytes": 1794048,
"scriptDurationMs": 3022.6169999999997,
"eventListeners": -174,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.80000000000109
},
{
"name": "vue-large-graph-idle",
"durationMs": 16780.888000000006,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 16238.941,
"heapDeltaBytes": -68790600,
"heapUsedBytes": 171672328,
"domNodes": -8061,
"jsHeapTotalBytes": -18243584,
"scriptDurationMs": 249.583,
"eventListeners": -14928,
"totalBlockingTimeMs": 0,
"frameDurationMs": 18.333333333333332,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-idle",
"durationMs": 16506.599999999933,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 15998.425000000001,
"heapDeltaBytes": 10485964,
"heapUsedBytes": 184700124,
"domNodes": -8061,
"jsHeapTotalBytes": -16252928,
"scriptDurationMs": 256.89399999999995,
"eventListeners": -14928,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.80000000000291
},
{
"name": "vue-large-graph-pan",
"durationMs": 20439.01900000003,
"styleRecalcs": 158,
"styleRecalcDurationMs": 17.957999999999974,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 20186.864,
"heapDeltaBytes": -38785244,
"heapUsedBytes": 181742180,
"domNodes": -8061,
"jsHeapTotalBytes": -17694720,
"scriptDurationMs": 908.5190000000001,
"eventListeners": -14928,
"totalBlockingTimeMs": 72,
"frameDurationMs": 18.886666666666738,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "vue-large-graph-pan",
"durationMs": 20495.452999999998,
"styleRecalcs": 170,
"styleRecalcDurationMs": 17.788000000000025,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 20148.822000000004,
"heapDeltaBytes": -33261328,
"heapUsedBytes": 192369056,
"domNodes": -8061,
"jsHeapTotalBytes": -16068608,
"scriptDurationMs": 916.5179999999999,
"eventListeners": -14928,
"totalBlockingTimeMs": 49,
"frameDurationMs": 18.330000000000048,
"p95FrameDurationMs": 16.80000000000291
},
{
"name": "vue-node-resize-workload",
"durationMs": 6108.901999999944,
"styleRecalcs": 20,
"styleRecalcDurationMs": 78.08500000000002,
"layouts": 20,
"layoutDurationMs": 23.064,
"taskDurationMs": 6005.946000000001,
"heapDeltaBytes": -41088176,
"heapUsedBytes": 184826472,
"domNodes": -8061,
"jsHeapTotalBytes": 15912960,
"scriptDurationMs": 105.19000000000001,
"eventListeners": -14928,
"totalBlockingTimeMs": 418,
"frameDurationMs": 17.779999999999927,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-node-resize-workload",
"durationMs": 6039.664000000016,
"styleRecalcs": 20,
"styleRecalcDurationMs": 77.38899999999998,
"layouts": 20,
"layoutDurationMs": 22.93200000000001,
"taskDurationMs": 5958.423000000001,
"heapDeltaBytes": -35705332,
"heapUsedBytes": 184669988,
"domNodes": -8061,
"jsHeapTotalBytes": 21512192,
"scriptDurationMs": 109.34299999999999,
"eventListeners": -14928,
"totalBlockingTimeMs": 369,
"frameDurationMs": 17.780000000000047,
"p95FrameDurationMs": 16.80000000000291
},
{
"name": "vue-zoom-out-idle",
"durationMs": 8273.319000000014,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 8046.294,
"heapDeltaBytes": -12059356,
"heapUsedBytes": 179985644,
"domNodes": 0,
"jsHeapTotalBytes": -14155776,
"scriptDurationMs": 148.83,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.219999999999953,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-zoom-out-idle",
"durationMs": 6377.011000000039,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 6093.454999999999,
"heapDeltaBytes": 16039116,
"heapUsedBytes": 201844564,
"domNodes": 0,
"jsHeapTotalBytes": -28835840,
"scriptDurationMs": 101.68899999999998,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "workflow-execution",
"durationMs": 104.57399999995687,
"styleRecalcs": 8,
"styleRecalcDurationMs": 14.854999999999999,
"layouts": 2,
"layoutDurationMs": 0.5970000000000001,
"taskDurationMs": 78.95200000000003,
"heapDeltaBytes": 3203728,
"heapUsedBytes": 68989504,
"domNodes": 109,
"jsHeapTotalBytes": 262144,
"scriptDurationMs": 8.387,
"eventListeners": 33,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "workflow-execution",
"durationMs": 126.05099999996128,
"styleRecalcs": 8,
"styleRecalcDurationMs": 20.043,
"layouts": 3,
"layoutDurationMs": 1.4329999999999998,
"taskDurationMs": 90.04900000000002,
"heapDeltaBytes": 3198580,
"heapUsedBytes": 69055508,
"domNodes": 109,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 9.668000000000001,
"eventListeners": 33,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
}
]
}🔬 E2E Coverage
Top 10 files by uncovered lines
|
Codecov Report✅ All modified and coverable lines are covered by tests. @@ Coverage Diff @@
## cloud/1.55 #19524 +/- ##
=============================================
Coverage ? 84.05%
=============================================
Files ? 3080
Lines ? 151859
Branches ? 39095
=============================================
Hits ? 127639
Misses ? 22811
Partials ? 1409
Flags with carried forward coverage won't be shown. Click here to find out more.
🚀 New features to boost your workflow:
|
7fbf0bb to
f46a2fa
Compare
This comment has been minimized.
This comment has been minimized.
2006485 to
9b4acad
Compare
c327dfc to
fc70f56
Compare
| 'processing_error', | ||
| 'payment_not_completed', | ||
| 'generic' | ||
| ] as const satisfies readonly NonNullable<OpStatus['decline_reason']>[] |
There was a problem hiding this comment.
Can we not just use this directly?
There was a problem hiding this comment.
Removed the copied decline-reason list in commit 9226516.
fc70f56 to
5c024a1
Compare
DrJKL
left a comment
There was a problem hiding this comment.
Verdict: changes requested. The head fails typecheck because the regenerated int64 fields reach account-core and the app as bigint, and billing-web handles payment_not_completed as a card decline.
Since my last review at fc70f56, the branch was rebuilt. The regen commit is now 25b0910, from cloud@b7db355, and it adds balance_applied_cents. The account-core commit 5c024a1 did not change (git range-diff shows =). I also missed the red typecheck when I approved fc70f56. It was failing there too, so that approval should not count.
Pre-merge blockers
- P1, inline on
zod.gen.ts:1049. The new preview money fields parse tobigint. The account-core adapter and its compile-time pins were not updated, sopnpm typecheckfails in account-core and in the app. - P1, inline on
zod.gen.ts:2556.amount_charged_centsandcredits_addedparse tobigintthroughBillingOpStatus, andcustomerActionParity.test.ts:43fails typecheck. - P1, body item 1 below. billing-web shows
payment_not_completedas a declined card and renders a missing translation key.
Findings without an anchorable line (3)
-
P1: billing-web classifies
payment_not_completedas a card decline and has no copy for it.apps/billing-web/src/checkout/payVerdict.ts:93-109routes a decline to thenot_completedcard only when the reason is inUNAUTHENTICATED_REASONS(authentication_failed,authentication_required) or isgenericwith a retry action. The new reason is in neither, andstepForReasoninpaymentProjection.tsmaps it todeclined. So a customer who abandoned approval gets the "declined" outcome, andInlineOutcomeCard.vue:50renderst('checkout.fullPage.outcome.reasons.payment_not_completed'). That key is missing fromapps/billing-web/src/locales/en/main.json:226-236, so vue-i18n prints the key path. I checked this with a throwaway billing-web test at this head:operationOutcomeOf(failedOperation('payment_not_completed', 'op_9'))returned{ kind: 'declined', reason: 'payment_not_completed', operationId: 'op_9' }.operationOutcomeOf(failedOperation('authentication_failed', 'op_8'))returned{ kind: 'not_completed', operationId: 'op_8' }.Object.keys(en.checkout.fullPage.outcome.reasons)does not containpayment_not_completed.
The server's own description ("customer non-approval or an expired attempt") matches billing-web's
not_completedcard. Adding'payment_not_completed'toUNAUTHENTICATED_REASONSfixes the card. Add a row for it in the existingpayVerdict.test.tsoperationOutcomeOftable next to theauthentication_failedrow. Also add thereasons.payment_not_completedlocale key so a futuredeclinedpath cannot print a raw key. -
P2: the
declineDetailKeytable inpaymentCopy.test.tshas no row for the new reason. Inpackages/account-core/src/core/billing/paymentCopy.test.ts:100-112every other decline reason has a row. I changed thepayment_not_completedcase inpaymentCopy.tsto returnpaymentDeclinedDetail, ranvitest run src/core/billing/paymentCopy, and all 16 tests still passed. So the mapping this commit adds is unpinned. One row fixes it:['payment_not_completed', 'authenticationFailedDetail'], or whatever the inline question onpaymentCopy.ts:96settles on. The newDEFAULT_PAYMENT_COPYkey needs no test, becauseRecord<PaymentCopyKey, string>already fails compilation without it. -
P3: the PR description does not mention the hand-authored commit. The body is the generator template.
5c024a1changes customer-facing copy and decline handling in account-core, and a reviewer reading "Automated Ingest API Type Update" would skip it. A line naming that commit, the new decline reason, and the account-core/billing-web follow-through would help.
My earlier findings (2, still open)
- Shouldn't this reference a generated type? (
billingOpStatus.contract.test.ts:35) is still open. The code did not change. I've refined it in the inline comment on line 114. - Can we not just use this directly? (
billingOpStatus.contract.test.ts:45) is also still open. Using the zod enum's.optionswould remove the copied literals. It would also leave theit.forchecking that a schema accepts its own options. Because production code already enforces exhaustiveness, I now think the table should go (see the line 114 comment).
Checked and cleared (0 rejected)
No candidate was rejected. I lowered two rows from the lane severity:
- The
paymentCopy.ts:96mapping went from P2 to a P3 question, because billing-web already groups authentication declines as "not completed", so the author's grouping is consistent. - The missing
declineDetailKeyrow went from P1 to P2. It is a regression-test gap, not a runtime defect.
Other checks:
git range-diff fc70f56~2..fc70f56 5c024a1~2..5c024a1shows the account-core commit unchanged. The regen delta is the b7db355 field additions only.pnpm exec eslintandoxfmt --checkpass on the two hand-edited files.- The generated files match the
openapi-ts.config.tsplugin set (@hey-api/typescriptpluszod, compatibility version 3). I could not regenerate them, becauseopenapi.yamlis not in the repository.
Coverage
- Depth: standard. Lanes run as separate reviewers: correctness, maintainability, idiomaticity, tests. I skipped documentation because the PR body is generator boilerplate with no behavior or compatibility claim, and one body item above covers it.
- Lenses (standard set): blast-radius, regression-risk-reviewer, idempotency, shared-state, boundary-discipline, fix-root-causes; ponytail, no-comments (applied by hand, no sub-dispatch available), minimize-reader-load, subtract-before-you-add, laziness-protocol, migrate-callers; typescript-best-practices, type-system-discipline; designing-test-suites, reviewing-unit-tests, test-mock-quality-analysis, consolidating-test-setup, prove-it-works.
- Checks at
5c024a1:pnpm typecheckexit 2 (8 errors insrc/platform/workspace/**).pnpm typecheck:account-coreexit 2 (3 pin errors).vitest run src/core/billingin account-core: 37 files, 551 passed. CItypecheck,typecheck-aux,lint-and-format, andverify-pack-contents (account-core)fail on this head and pass on base44cbbe35. - Nobody traced:
apps/website. It does not import the changed shapes.
| message: 'Invalid value: Expected int64 to be <= 9223372036854775807' | ||
| }) | ||
| .optional(), | ||
| balance_applied_cents: z.coerce |
There was a problem hiding this comment.
P1. This new int64 field, together with subtotal_cents (line 1113), PreviewPlanInfo.list_price_cents (line 1011), and SubscriptionDiscount.duration_in_months (line 554), parses to bigint, but types.gen.ts declares all four as number. Account-core's PreviewSchema / PlanInfoSchema in packages/account-core/src/core/billing/subscriptionCommands.ts:125-146 convert each money field to a safe number one by one, and nobody added these four. So SubscriptionPreview now carries bigint for them, and that breaks every place the app hands it back as a generated PreviewSubscribeResponse:
pnpm typecheckat this head fails insubscriptionOperationView.ts:199,useSubscriptionCheckout.ts:661and:715,SubscriptionAddPaymentPreviewWorkspace.vue:11,SubscriptionSuccessWorkspace.vue:6,SubscriptionTransitionPreviewWorkspace.vue:3, anduseWorkspaceBilling.subscriptionRail.test.ts:482.pnpm typecheck:account-corefails where the compile-time pins were built to catch exactly this.previewSubscribe.contract.test.ts:75wantsbalance_applied_centsandsubtotal_centsinMONEY_FIELDS.subscriptionCommands.test.ts:809wants them inQUOTE_CENT_FIELDS, and:812wantslist_price_centsinPLAN_CENT_FIELDS.- At runtime, once the server sends these fields,
createBillingCommands().previewSubscribereturnscost_today_centsas a number and the new fields as bigint.JSON.stringifyon the result throws "Do not know how to serialize a BigInt", and adding them to a number throws.
Smallest fix: add balance_applied_cents: cents.optional() and subtotal_cents: cents.optional() to PreviewSchema, list_price_cents: cents.optional() to PlanInfoSchema, and duration_in_months: z.number().int().safe().optional() to the discount extension. Then add the three _cents names to the three pin tables. The existing coercion, int64-range, and rejectsQuote rows then cover the new fields with no new test bodies.
There was a problem hiding this comment.
Normalized the new preview fields and extended the existing field-table coverage in commit 9965529.
| */ | ||
| export const zBillingOpStatusResponse = z.object({ | ||
| action_url: z.string().optional(), | ||
| amount_charged_cents: z.coerce |
There was a problem hiding this comment.
P1. amount_charged_cents (here) and credits_added (line 2575) parse to bigint, but BillingOpStatusResponse in types.gen.ts types them as number. Account-core derives BillingOpStatus straight from zBillingOpStatusResponse (operationState.ts:19) with no number adapter. So a polled status now carries bigint, and src/platform/workspace/billing/customerActionParity.test.ts:43 fails typecheck: status.amount_charged_cents is number | undefined against bigint | undefined.
The server only sends these on succeeded operations, and they are display-only. Nothing reads them yet, so the runtime risk today is serialization: a status that reaches JSON.stringify (logging, persistence, cross-tab messaging) throws. I'd normalize them where the preview does. Extend the status schema with amount_charged_cents: cents.optional() and credits_added: z.number().int().safe().optional(), and derive BillingOpStatus from that, so the parity fixture typechecks without changes.
There was a problem hiding this comment.
Normalized operation status amounts through the Account Core schema in commit 9965529.
| * What a credit top-up of the requested amount would grant. | ||
| */ | ||
| export const zTopupQuoteResponse = z.object({ | ||
| amount_cents: z.coerce |
There was a problem hiding this comment.
P3. Every int64 field is emitted twice with different types: number in types.gen.ts, z.coerce.bigint() here. TopupQuoteResponse.amount_cents / credits is the newest case. It has no consumer yet, so the first caller of zCreateTopupQuoteResponse gets bigint where the generated type says number. This PR adds seven more fields with the split, and each one needs a hand-written adapter in account-core, which is how the two P1s above happened. Not a blocker here, but it would be worth a follow-up to set the int64 handling in packages/ingest-types/openapi-ts.config.ts, so the Zod output matches the TS output and account-core can drop its per-field cents overrides.
There was a problem hiding this comment.
Filed #19704 to align generated Zod int64 output with the generated TypeScript number fields.
| return 'incorrectCvcDetail' | ||
| case 'authentication_required': | ||
| case 'authentication_failed': | ||
| case 'payment_not_completed': |
There was a problem hiding this comment.
P3, question. The generated doc defines payment_not_completed as "Stripe explicitly reported customer non-approval or an expired attempt". This routes it to authenticationFailedDetail ("We couldn't complete payment verification. Please try again."). That fits an abandoned 3DS or wallet approval, and it matches billing-web, which counts authentication declines as "not completed". It reads oddly for an attempt that simply expired. Is "verification" the wording you want, or should this get its own detail key? Either way, please pin the answer with a row in the declineDetailKey table in paymentCopy.test.ts (see the review body).
There was a problem hiding this comment.
| } | ||
| ) | ||
|
|
||
| it.for(DECLINE_REASONS)( |
There was a problem hiding this comment.
P3. This carries forward my two earlier comments on lines 35 and 45. I'd delete DECLINE_REASONS, its toEqualTypeOf pin on line 57, and this it.for, instead of deriving the list from the generated enum:
- The rows come from the enum that
zBillingOpStatusResponsedefines, and the test asserts that the same schema accepts them. No wrong implementation fails here that the compiler doesn't already catch. - Production already makes a regen fail when a reason is added. Removing the
payment_not_completedcase fromdeclineDetailKeyfails account-core typecheck with TS2366 atpaymentCopy.ts:86. Removing itsDEFAULT_PAYMENT_COPYentry fails with TS2741 atpaymentCopy.ts:29. - Switching to
zBillingOpStatusResponse.shape.decline_reason.unwrap().optionswould remove the copied literals but make the test fully tautological.
Deleting these 28 lines adds no new typecheck error. The regeneration guard stays in paymentCopy.ts, and the behavior guard is the declineDetailKey table row I asked for in the body.
There was a problem hiding this comment.
Removed the redundant enum pin and tautological schema loop in commit 9226516.
5c024a1 to
0d6a4e0
Compare
Addresses review feedback: #19524 (comment) #19524 (comment) #19524 (comment)
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
🔧 Auto-fixes AppliedThis PR has been automatically updated to fix linting and formatting issues.
Changes made:
|
DrJKL
left a comment
There was a problem hiding this comment.
Verdict: one P2 left before approval. The charge-breakdown blocker is fixed and pnpm typecheck is green, but the new shared centsSchema coerces, so every money field now accepts null, "", and booleans as amounts.
Since my last review at b89d5e9:
- The branch was rebuilt on cloud/1.55 at
e9a89db. The new base commits are agent backports (#19732, #19803). - The six earlier commits are unchanged (
git range-diffshows=). 7f25acbaddsmoneySchemas.ts, normalizescharge_breakdownthrough it, movesPreviewSchema/PlanInfoSchemaonto the shared schemas, and adds a breakdown test.8e510b6is the formatter.
The author already replied in the charge-breakdown thread, so I posted no thread reply this round.
Remaining
- P2, inline on
moneySchemas.ts:4.z.coerce.number()turnsnull,""," ",false, and[]into0, andtrueinto1, for every preview and op-status amount, including required ones likecost_today_cents. The previousz.number()rejected all of them. The coercion is only needed because the new test feedsbigintliterals: reverting it fails that one test and nothing else. - P3 Nit, inline on
subscriptionCommands.ts:123. The int64 rationale docblock now sits abovePlanInfoSchemainstead of thecentsdefinition it explains.
Earlier findings (all from prior rounds)
- P1
charge_breakdownint64 amounts reachedBillingOpStatusasbigint. Addressed in7f25acb. Author reply.pnpm typechecknow exits 0, andcustomerActionParity.test.ts:43is clean. - P1
monthly_list_price_cents/monthly_price_cents. Addressed ind16cbd9, carried as58b3848. - P1 required placement key broke
RemoteConfig. Addressed inb89d5e9, carried as14a1957. - P1 preview, op-status, and proration int64 fields (r4148675543, r4148675566, r4151572288). Still fixed, now through
centsSchema. - P3 op-status normalization test. Still addressed (
7fb1017), and7f25acbextends it to the breakdown. - P3
DECLINE_REASONSdeletion (r4147777000, r4147782987, r4148675608). Still addressed (69915bc). - P3 generator int64 split. Deferred to #19704.
- P3 strict parse of the display-only receipt fields. Coercion happens to let
"123"through now, but by acceptingnullas0, which is the wrong trade. If the P2 is reverted, this stays an open, non-blocking P3. - P3 Nit:
previewSubscribe.contract.test.ts:76still says "a seventh amount". Open. - Round 2's
paymentCopy.ts:96question, billing-web decline handling, missingdeclineDetailKeyrow, and PR-description note were superseded by #19670 or the dropped commit, as recorded in my round 4 review.
Coverage
Depth: standard, delta-scoped to 7f25acb and 8e510b6. The six earlier commits are byte-identical to the ones I reviewed at b89d5e9. Checks at 8e510b6:
pnpm typecheckexit 0 (app, account-core, billing-web, account-ui, billing-contract).vitest run src/core/billingin account-core: 566 passed. WithcentsSchemareverted toz.number().int().safe(), 1 failed (normalizes charge breakdown amounts to numbers) and 565 passed.- Scratch probe of
centsSchemaagainstz.number().int().safe()overnull,""," ","abc","12",true,false,[],12.5,1999,1999n: results in the inline table. - ESLint on the four changed account-core files exit 0.
- CI:
verify-pack-contents (account-core)passed.typecheckandtypecheck-auxwere still running when I posted.
| import { zSubscriptionDiscount } from '@comfyorg/ingest-types/zod' | ||
| import { z } from 'zod' | ||
|
|
||
| export const centsSchema = z.coerce.number().int().safe() |
There was a problem hiding this comment.
P2. Switching to z.coerce.number() loosens every money field account-core parses, not just the new breakdown. Coercion runs Number(x) before .int().safe(), so values that used to be MALFORMED_RESPONSE now parse as amounts. I ran both schemas at this head (scratch test, deleted after):
| input | z.coerce.number().int().safe() |
previous z.number().int().safe() |
|---|---|---|
null |
0 |
rejected |
"" / " " |
0 |
rejected |
false / true |
0 / 1 |
rejected |
[] |
0 |
rejected |
"12" |
12 |
rejected |
1999n |
1999 |
rejected |
That includes required quote fields like cost_today_cents, which the docblock in subscriptionCommands.ts describes as "confirmed as charges". A null there now renders as a $0.00 charge instead of failing the quote.
Coercion seems to exist only because the new breakdown test feeds bigint literals (1999n, 500n, 3n), which never appear in a JSON body. With this line set back to z.number().int().safe(), that test is the only one of 566 that fails. The rest of src/core/billing passes. I'd revert the coercion here and on line 8, and give the breakdown test plain numbers. If the OpStatusBody fixture type objects to numbers there, type that one body as z.input<typeof BillingOpStatusSchema>. That is the shape the lifecycle actually parses, so no cast is needed.
There was a problem hiding this comment.
Fixed in 9629acb. Money schemas are strict again, and the breakdown regression now uses JSON number inputs.
| */ | ||
| const cents = z.number().int().safe() | ||
|
|
||
| const PlanInfoSchema = zPreviewPlanInfo.extend({ |
There was a problem hiding this comment.
P3, Nit. This docblock explained the const cents that used to sit right below it. With cents moved to moneySchemas.ts, it now reads as documentation for PlanInfoSchema, and "so the cents are read as numbers" no longer points at anything. Moving it onto centsSchema in moneySchemas.ts keeps the rationale with the one definition both schemas now share.
There was a problem hiding this comment.
Moved the rationale beside centsSchema in 2b5653a.
🔧 Auto-fixes AppliedThis PR has been automatically updated to fix linting and formatting issues.
Changes made:
|
DrJKL
left a comment
There was a problem hiding this comment.
Verdict: one P2 left before approval. The money-parsing and docblock findings are fixed, but this head adds an unrelated #19635 cherry-pick whose AI co-author trailer fails the trailer check.
Since my last review at 8e510b6, three commits landed on the same base, e9a89db:
9629acbsetscentsSchemaandduration_in_monthsback toz.number().int().safe()and feeds the breakdown test JSON numbers. The author replied.2b5653amoves the int64 rationale docblock ontocentsSchema. The author replied.b464087cherry-picks #19635, which deletes a live-fetch test fromapps/website/src/config/router-providers.test.ts.
The author answered both of my threads, so I posted no thread replies.
Remaining
- P2, inline on the deleted
router-providers.test.tshunk.b464087is unrelated to this PR and carriesCo-authored-by: Cursor Agent.Check for AI agent co-author trailersfails on this head. Drop it here and backport #19635 separately without the trailer.
The account-core and remote-config side is done. Once b464087 is out and CI is green, I have nothing left that blocks.
Findings and where they stand
- P2
z.coerce.number()acceptednull/""/ booleans as amounts. Addressed in9629acb. With thecharge_breakdownoverride removed, the rewritten breakdown test fails with"amount_charged_cents": 1999n, so it still pins the fix with realistic JSON input. - P3 Nit orphaned int64 rationale docblock. Addressed in
2b5653a. - P1
charge_breakdownint64 amounts. Addressed in7f25acb. - P1
monthly_list_price_cents/monthly_price_cents. Addressed in58b3848. - P1 required placement key broke
RemoteConfig. Addressed in14a1957. - P1 preview, op-status, and proration int64 fields (r4148675543, r4148675566, r4151572288). Addressed in
9c0ba0c, now throughcentsSchema. - P3 op-status normalization test. Addressed in
7fb1017. - P3
DECLINE_REASONSdeletion (r4147777000, r4147782987, r4148675608). Addressed in69915bc. - P3 generator int64 split. Deferred to #19704.
- P3 strict parse of the display-only receipt fields. Open, not blocking, and I'm fine leaving it as is. With strict parsing restored, this is the trade-off I described originally.
- P3 Nit:
previewSubscribe.contract.test.ts:76still says "a seventh amount". Open, not blocking. - Round 2's
paymentCopy.ts:96question, billing-web decline handling, missingdeclineDetailKeyrow, and PR-description note were superseded by #19670 or the dropped commit (round 4 review).
Coverage
Depth: standard, delta-scoped to 9629acb, 2b5653a, and b464087. Checks at b464087:
pnpm typecheckexit 0.vitest run src/core/billingin account-core: 566 passed. With thecharge_breakdownoverride removed, the breakdown row fails (1999n).git log origin/cloud/1.55..HEAD:b464087is the only commit with aCo-authored-bytrailer. #19635 merged tomainon 2026-09-30 and is not on cloud/1.55.- CI:
Check for AI agent co-author trailersfailed.typecheck-auxandverify-pack-contents (account-core)passed.typecheckandwebsite-unitwere still running when I posted.
| ) | ||
| }) | ||
|
|
||
| it('counts the models the docs catalog lists', async (ctx) => { |
There was a problem hiding this comment.
P2. This deletion arrives through b464087, a cherry-pick of #19635, which has nothing to do with the ingest types. It also brings over the original commit's Co-authored-by: Cursor Agent <cursoragent@cursor.com> trailer. The repo's Check for AI agent co-author trailers job fails on this head for that reason ("AI agent Co-authored-by trailers detected in PR commits"), and b464087 is the only commit here with a Co-authored-by line. apps/website/AGENTS.md describes that check as failing the whole PR.
I'd drop b464087 from this branch and backport #19635 to cloud/1.55 as its own PR, with the trailer removed from the cherry-pick message. That keeps this PR to the regen plus its account-core and remote-config follow-through. It also means the next regen push doesn't have to replay an unrelated website change.
There was a problem hiding this comment.
Dropped the unrelated #19635 cherry-pick. Current head f1969adad279 contains the ingest update and the capability fixture repairs required by its generated type changes.
DrJKL
left a comment
There was a problem hiding this comment.
Verdict: changes requested. The #19635 cherry-pick is gone and the trailer check passes, but the cloud@44b2ac7 regen adds a required can_revert_scheduled_change capability: four builders weren't updated, so account-core and billing-web typecheck are red. The runtime parse also needs a deploy-order answer.
Since my last review at b464087:
- The branch was rebuilt on cloud/1.55 at
f1b9f07; the new base commits are agent backports #19832, #19727, and #19733. - The ten account-core and remote-config commits are unchanged (
git range-diffshows=). b464087is dropped, so this addresses my trailer P2, which the author already answered in the thread.- The regen,
91cdce7, adds:- the revert-scheduled-change endpoint and
BillingCapabilities.can_revert_scheduled_change, 'cancelled'on theTaskEntry/TaskResponse/AssetDownloadResponsestatuses,CancelTask*.
- the revert-scheduled-change endpoint and
62c8ed4is the formatter.f1969adadds the new key to five capability fixtures.
Pre-merge blockers
- P1, inline on
zod.gen.ts:2812. Four capability builders still lackcan_revert_scheduled_change:billingCapabilities.contract.test.ts:17and:55, billing-webe2e/fixtures/scenario.ts:36, andsrc/test/fakeBillingClient.ts:186.typecheck:account-coreandtypecheck:billing-webexit 2. CItypecheckandverify-pack-contents (account-core)fail. - P1 question, inline on
types.gen.ts:4587.capabilities.tsparses capabilities with the generated schema, so a backend that doesn't send the new key makes every capabilities read malformed. Please confirm the deploy order, or default the key tofalseinCapabilitiesBodySchema.
Earlier findings
- P2 unrelated #19635 cherry-pick with an AI co-author trailer. Addressed: dropped at this head, and
Check for AI agent co-author trailerspasses. Author reply. - P2
z.coerce.number()incentsSchema. Still fixed (664ad6e). - P3 Nit orphaned rationale docblock. Still fixed (
ea3ad4c). - P1
charge_breakdownint64 amounts, monthly plan prices,RemoteConfigplacement key, and the preview, op-status, and proration fields (r4148675543, r4148675566, r4151572288). Still fixed at this head.typecheck:appexits 0, and account-core billing tests pass. - P3 op-status normalization test and the
DECLINE_REASONSdeletion (r4147777000, r4147782987, r4148675608). Still addressed. - P3 generator int64 split. Deferred to #19704. This regen adds no new int64 fields.
- P3 strict parse of the display-only receipt fields. Open, not blocking.
- P3 Nit:
previewSubscribe.contract.test.ts:76still says "a seventh amount". Open. - Round 2's
paymentCopy.ts:96question, billing-web decline handling, missingdeclineDetailKeyrow, and PR-description note were superseded by #19670 or the dropped commit (round 4 review).
Checked and cleared
- The new
'cancelled'task status: nothing insrc/,packages/, orapps/consumesTaskEntry,TaskResponse, orAssetDownloadResponseoutside the generated package, so no switch or mapping misses the new member. RevertScheduledChange*andCancelTask*are new endpoint types with no consumer yet.
Coverage
Depth: standard, delta-scoped to the regen, 62c8ed4, and f1969ad. Checks at f1969ad:
pnpm typecheck:appexit 0.pnpm typecheck:account-coreexit 2 (TS2741 at:17, TS2344 at:55).pnpm typecheck:billing-webexit 2 (2 errors).typecheck:account-ui,typecheck:billing-contract, andtypecheck:browserexit 0.zBillingCapabilities.safeParseon the pre-regen capabilities object:success=false,can_revert_scheduled_change: Required.- CI:
typecheckandverify-pack-contents (account-core)failed.typecheck-auxandCheck for AI agent co-author trailerspassed.website-unitwas still running when I posted.
| can_downgrade_to_personal: z.boolean(), | ||
| can_invite_members: z.boolean(), | ||
| can_reactivate: z.boolean(), | ||
| can_revert_scheduled_change: z.boolean(), |
There was a problem hiding this comment.
P1. The cloud@44b2ac7 regen adds a required can_revert_scheduled_change: boolean to BillingCapabilities. f1969ad updates five fixtures, but four more places still build a capabilities object without the key, so typecheck is red at this head:
packages/account-core/src/core/billing/contracts/billingCapabilities.contract.test.ts:17, thecapabilities()builder: TS2741,Property 'can_revert_scheduled_change' is missing.- The same file at
:55, theexpectTypeOf<Capabilities>().toEqualTypeOf<{ ... }>()pin: TS2344. The pin is supposed to fail on a regen like this; it just needs the new key. apps/billing-web/e2e/fixtures/scenario.ts:36(capabilitiesWith):can_revert_scheduled_change?: boolean | undefinedis not assignable toBillingCapabilities.apps/billing-web/src/test/fakeBillingClient.ts:186(capabilitiesSnapshot): same error.
pnpm typecheck:account-core and pnpm typecheck:billing-web both exit 2. CI typecheck and verify-pack-contents (account-core) fail on the first two. typecheck:app, typecheck:browser, typecheck:account-ui, and typecheck:billing-contract pass. Adding can_revert_scheduled_change: false to each of the three builders and to the pin clears them.
There was a problem hiding this comment.
Fixed in fc67699. The remaining capability builders and compile-time pin now include can_revert_scheduled_change.
| /** | ||
| * Stripe-billed only; false within 1 hour of the change. | ||
| */ | ||
| can_revert_scheduled_change: boolean |
There was a problem hiding this comment.
P1, question on deploy order. This field is also required at runtime, not just in types. capabilities.ts parses GET capabilities through CapabilitiesBodySchema, which embeds the generated zBillingCapabilities. Against a body without the key, zBillingCapabilities.safeParse returns success=false with [["can_revert_scheduled_change","Required"]] (checked at this head). So if cloud/1.55 reaches staging or prod before the backend that emits this field, every capabilities read comes back malformed. Every billing gate that reads capabilities then loses its snapshot, not just the new revert action. That is the same ordering hazard #19670 described for decline_reason, and the field only landed in cloud between 2e53e9f and 44b2ac7.
Is the backend that sends can_revert_scheduled_change already deployed everywhere this frontend will ship? If it isn't, or nobody is sure, I'd make the reader tolerant for this one key: zBillingCapabilities.extend({ can_revert_scheduled_change: z.boolean().default(false) }) in CapabilitiesBodySchema. That fails closed for the new action only, which fits the schema's own "conservative UI guidance" description, and BillingCapabilities stays boolean for consumers.
There was a problem hiding this comment.
Applied the fail-closed compatibility path in 7d8ed09. Older responses now default only can_revert_scheduled_change to false; focused capability tests pass 33/33.
Addresses review feedback: #19524 (comment) #19524 (comment) #19524 (comment)
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
Fixes deterministic typecheck failures: https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/36954806939/job/110675299405
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
Fixes the typecheck failure reported by CI: https://github.com/Comfy-Org/ComfyUI_frontend/actions/runs/37074360738/job/111060863559
DrJKL
left a comment
There was a problem hiding this comment.
Verdict: one P2 left before approval. Both capability findings are fixed and typecheck is green, but the rebase typed #19830's raw placement flag as the generated enum, and the follow-up test fix gets past that type with Reflect.set.
Since my last review at f1969ad:
- The branch was rebuilt on cloud/1.55 at
40a0284, ten base commits later. One of them is #19830, the free-use placement experiment. git range-diffshows every replayed commit unchanged exceptb2c89df(theRemoteConfigchange). Its conflict resolution dropped #19830's line.- Three commits are new:
The author answered both of my capability threads, so I posted no thread replies.
Remaining
- P2, inline on
types.ts:89andfreeUsePlacement.test.ts:43.RemoteConfigtypes'agent-free-use-message-placement'as the generated enum, but the config is stored without validation, and #19830'sisFreeUseVariantexists for exactly the unknown values the type rules out. Restore the rawstringline, drop thePartial<Pick<...>>, and revertdcdd347. I verified that this combination typechecks.
Earlier findings
- P1 four capability builders lacked
can_revert_scheduled_change. Addressed infc67699.typecheck:account-coreandtypecheck:billing-webexit 0. - P1 question required capability key versus older backends. Addressed in
7d8ed09. Only the new key defaults tofalse, andBillingCapabilitiesstaysboolean. The new assertion incapabilities.test.tsruns against aCAPABILITIESfixture without the key, so it exercises the default. - P2 unrelated #19635 cherry-pick. Still gone. The trailer check passes.
- P2
centsSchemacoercion and P3 docblock placement. Still fixed. - P1 int64 normalization: charge breakdown, monthly prices, and the preview, op-status, and proration fields (r4148675543, r4148675566, r4151572288). Still fixed.
- P1 required placement key broke
RemoteConfig. The original fix still holds: the key is optional and the other fields keep their contract types. The P2 above refines it now that #19830 consumes the key as raw data. - P3 op-status normalization test and the
DECLINE_REASONSdeletion (r4147777000, r4147782987, r4148675608). Still addressed. - P3 generator int64 split. Deferred to #19704.
- P3 strict parse of the display-only receipt fields and the P3 Nit on
previewSubscribe.contract.test.ts:76("a seventh amount"). Open, not blocking. - Round 2's
paymentCopy.ts:96question, billing-web decline handling, missingdeclineDetailKeyrow, and PR-description note were superseded by #19670 or the dropped commit (round 4 review).
Coverage
Depth: standard, delta-scoped to fc67699, 7d8ed09, dcdd347, and the changed replay of b2c89df against the new base. Checks:
- At
7d8ed09:pnpm typecheck:appexit 2 atfreeUsePlacement.test.ts:44. With the suggestedtypes.tschange applied locally, exit 0. - At
dcdd347:pnpm typecheck:appexit 0.typecheck:account-core,typecheck:billing-web,typecheck:account-ui,typecheck:billing-contract, andtypecheck:browserexit 0 (run at7d8ed09, anddcdd347changes only the test). vitest run src/core/billingin account-core: 566 passed.freeUsePlacementtests: 4 passed.- ESLint on
freeUsePlacement.test.tsis clean. - CI at
dcdd347:typecheck,typecheck-aux,verify-pack-contents (account-core), andCheck for AI agent co-author trailerspassed, and the rest were still running when I posted.
| [K in keyof FeaturesResponse as K extends PlacementKey | ||
| ? never | ||
| : K]: FeaturesResponse[K] | ||
| } & Partial<Pick<FeaturesResponse, PlacementKey>> & { |
There was a problem hiding this comment.
P2. The rebase onto cloud/1.55 dropped a field the base had just added. #19830 (80673ee) declared 'agent-free-use-message-placement'?: string in this literal. That is the raw wire value, which useFreeUsePlacement checks with isFreeUseVariant and treats as 'control' when it's unknown. The replayed b2c89df removed that line, so the key now comes from Partial<Pick<FeaturesResponse, PlacementKey>> typed as the generated five-arm enum.
The value is never validated at the boundary. commitRemoteConfigSuccess stores the response as-is, so an unknown string can reach remoteConfig while the type says it can't. That is why #19830's it.for([undefined, 'unknown'])('keeps control for %s') row stopped compiling at 7d8ed09, and why dcdd347 now writes it with Reflect.set (see the comment on that line). The honest type is the raw string this literal already uses for hosted_billing_destination and signup_turnstile.
Fix: keep the mapped type that drops the key, remove the Partial<Pick<...>> part, and put #19830's line back:
} & {
...
'agent-in-app-experience'?: boolean
'agent-free-use-message-placement'?: stringI applied that at 7d8ed09, with the test still using the plain remoteConfig.value = { ... } assignment, and pnpm typecheck:app exited 0. Then dcdd347 can be reverted. This is a deliberate exception to the "don't re-declare an omitted generated field" rule in typescript.md. The field is unvalidated wire data with a validator downstream, so typing it as the contract enum hides the cases that validator handles.
There was a problem hiding this comment.
Fixed in 84c05b3: the placement field is raw optional string data again, while the mapped generated response still excludes its narrower enum type. App typecheck passes.
| remoteConfig.value = { | ||
| 'agent-free-use-message-placement': value | ||
| } | ||
| Reflect.set(remoteConfig.value, 'agent-free-use-message-placement', value) |
There was a problem hiding this comment.
P2, same root cause as my comment on src/platform/remoteConfig/types.ts:89. Reflect.set writes a value the declared type forbids, which works the same as a cast. AGENTS.md and docs/guidance/typescript.md ask tests to model malformed wire input as data rather than bypass the type to get it in. The test has to do this only because RemoteConfig now types the field as the generated enum. Once the field is back to string, the original remoteConfig.value = { 'agent-free-use-message-placement': value } compiles again, and this commit can be reverted.
There was a problem hiding this comment.
Fixed in 42e61a8: the test models the unknown wire value with the typed remote-config assignment again; the focused suite passes 4/4.
Addresses review feedback: #19524 (comment)
Addresses review feedback: #19524 (comment)
DrJKL
left a comment
There was a problem hiding this comment.
Verdict: ready to approve. 84c05b3 puts #19830's raw 'agent-free-use-message-placement'?: string back in RemoteConfig, and 42e61a8 reverts the Reflect.set workaround. Every typecheck is green. Two P3 items stay open, and neither blocks.
The author answered both threads from my last review: types.ts and the test. So I posted no thread replies. The mapped type still drops the generated enum for that key, so the placement value has a single type: the raw string isFreeUseVariant checks. freeUsePlacement.test.ts is now identical to cloud/1.55, so this PR no longer touches it.
Still open, not blocking:
- P3 strict parse of the display-only receipt fields. I'm fine leaving it as is.
- P3 Nit:
previewSubscribe.contract.test.ts:76still says "so a seventh amount reaches both the coercion rows and the int64 rows", andMONEY_FIELDShas ten entries.
Review record
Commit SHAs below are the ones on the current head. The branch was rebased and replayed several times, so earlier rounds cited other SHAs for the same commits.
Round 1, comments and approval at fc70f56:
- P3
DECLINE_REASONSshould reference a generated type. Addressed in977c53f(list deleted). My reply. - P3 use the generated enum directly. Addressed in
977c53f. Author reply. - The approval itself: withdrawn in round 2, because typecheck was already red at
fc70f56.
Round 2, changes requested at 5c024a1:
- P1 preview int64 fields reached
SubscriptionPreviewasbigint. Addressed in523d5dc. Author reply. - P1 op-status
amount_charged_cents/credits_addedparsed tobigint. Addressed in523d5dc(BillingOpStatusSchema). Author reply. - P3 generator emits int64 as
numberin TS andbigintin Zod. Deferred to #19704. Author reply. - P3 mapping question on
paymentCopy.ts:96. Superseded when the hand commit was dropped and #19670 landed the mapping with a test row on base. My reply. - P3 delete the
DECLINE_REASONStable, pin, andit.for. Addressed in977c53f. Author reply. - P1 (body) billing-web showed
payment_not_completedas a card decline. Superseded by #19670 (payVerdict.ts:97). - P2 (body) missing
declineDetailKeyrow. Superseded by #19670 (paymentCopy.test.ts:123). - P3 (body) PR description omitted the hand commit. Superseded when that commit was dropped.
Round 3, changes requested at 2ebe00f:
- P1
proration_remaining_cents/proration_unused_centsparsed tobigint. Addressed in523d5dc. Author reply.
Round 4, ready to approve at 9226516:
- P3 no account-core test pinned op-status normalization. Addressed in
268a792. Author reply. - P3 strict parse of the display-only receipt fields. Open, left to the author.
- P3 Nit (body) stale "seventh amount" comment. Open.
Round 6, ready to approve at c8fdd78: no new findings.
Round 7, changes requested at 3a62316:
- P1
monthly_list_price_cents/monthly_price_centsparsed tobigint. Addressed inadc971a. My reply.
Round 9, changes requested at eb145b4:
- P1 required placement key broke
RemoteConfig. Addressed inb2c89df(key-remapped type), refined in84c05b3(raw string after #19830). My reply. - Contested: christian-byrne's approval claimed the generated TS and Zod shapes agreed while typecheck was red. Every point I raised then has since been fixed.
Round 10, changes requested at b89d5e9:
- P1
charge_breakdownint64 amounts parsed tobigint. Addressed in21f070a. Author reply.
Round 11, one P2 left at 8e510b6:
- P2
z.coerce.number()acceptednull/""/ booleans as amounts. Addressed in7d5b951. Author reply. - P3 Nit orphaned int64 rationale docblock. Addressed in
a874027. Author reply.
Round 12, one P2 left at b464087:
- P2 unrelated #19635 cherry-pick with an AI co-author trailer. Addressed: dropped. Author reply.
Round 13, changes requested at f1969ad:
- P1 four capability builders lacked
can_revert_scheduled_change. Addressed infc67699. Author reply. - P1 question required capability key versus older backends. Addressed in
7d8ed09, which defaults only that key tofalseand adds a test. Author reply.
Round 14, one P2 left at dcdd347:
- P2 rebase typed #19830's raw placement flag as the generated enum. Addressed in
84c05b3. Author reply. - P2
Reflect.setbypassed the type infreeUsePlacement.test.ts. Addressed in42e61a8(reverted to the plain assignment). Author reply.
Rounds 5 and 8 were replays with no change, so I posted nothing.
Checks at 42e61a8:
pnpm typecheckexit 0 (app, account-core, billing-web, account-ui, billing-contract), andpnpm typecheck:browserexit 0.vitest run src/core/billingin account-core: 566 passed.freeUsePlacementandsrc/platform/remoteConfig: 2 files, 28 passed.- ESLint and
oxfmt --checkonsrc/platform/remoteConfig/types.tsare clean. - CI:
typecheck,typecheck-aux,lint-and-format,verify-pack-contents (account-core),billing-web-unit-gate, andCheck for AI agent co-author trailerssucceeded, with no failures so far.testand 15 other jobs were still running when I posted.
christian-byrne
left a comment
There was a problem hiding this comment.
Approved on exact head 42e61a8 after self-review, DrJKL’s final ready-to-approve review, and final required CI/browser rollup completion. The two documented P3 follow-ups are non-blocking.
Automated Ingest API Type Update
This PR updates the Ingest API TypeScript types and Zod schemas from the latest cloud OpenAPI specification.
These types cover the FE-facing ingest API (workspaces, billing, secrets, assets, tasks, etc.).
Cloud's internal / machine-to-machine surface —
x-internaloperations and the/admin/,/api/internal/,/api/webhooks/rails — is stripped before generation(BE-2669), so it is not code-generated into this public package.