Repository navigation
fix(telemetry): report each assertion failure to RUM exactly once - #16333
Conversation
RUM collects console.error on its own, so every assertion that reportAssertFailure also reports arrived twice: once untagged from the console, once tagged. The untagged copy also bypassed the reporter's dedup and 20-per-session cap, so a render-loop invariant flooded RUM regardless of either safeguard. Drop the console-sourced copy, but only while a reporter exists to produce the tagged one, so a boot-path assertion is never silenced entirely.
…tay bounded reportAssertFailure dedups on exact message and caps a session at 20 distinct ones. Interpolating node ids into the message made every node its own message, so clearing a graph under identity drift burned the whole session budget in one loop — after which an assertion produced no tagged report and, with the console echo now dropped, no RUM event at all. Static messages keep cardinality at roughly one per call site, which is what the assert() docstring already asked for.
…riant message readSchemaVersion returns unknown from a merged Yjs doc, so the version is peer influenced and unbounded — the one assertion message that could still exhaust reportAssertFailure's session cap. FollowerSchemaError already carries `found` as a structured field, so the detail survives where it belongs.
🎭 Playwright: ✅ 1889 passed, 0 failed · 2 flaky📊 Browser Reports
🎨 Storybook: ✅ Built — View Storybook📦 Bundle: 9.66 MB gzip 🟢 -86 BDetailsSummary
Category Glance App Entry Points — 3.78 kB (baseline 3.78 kB) • ⚪ 0 BMain entry bundles and manifests
Status: 1 added / 1 removed Graph Workspace — 1.4 MB (baseline 1.4 MB) • ⚪ 0 BGraph editor runtime, canvas, workflow orchestration
Status: 2 added / 2 removed / 1 unchanged Views & Navigation — 124 kB (baseline 124 kB) • ⚪ 0 BTop-level views, pages, and routed surfaces
Status: 13 added / 13 removed / 4 unchanged Panels & Settings — 621 kB (baseline 621 kB) • ⚪ 0 BConfiguration panels, inspectors, and settings screens
Status: 10 added / 10 removed / 17 unchanged User & Accounts — 27.5 kB (baseline 27.5 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 — 115 kB (baseline 115 kB) • ⚪ 0 BReusable component library chunks
Status: 7 added / 7 removed / 8 unchanged Data & Services — 3.58 MB (baseline 3.58 MB) • 🟢 -138 BStores, services, APIs, and repositories
Status: 14 added / 14 removed / 3 unchanged Utilities & Hooks — 555 kB (baseline 555 kB) • ⚪ 0 BHelpers, composables, and utility bundles
Status: 17 added / 17 removed / 19 unchanged Vendor & Third-Party — 18.1 MB (baseline 18.1 MB) • ⚪ 0 BExternal libraries and shared vendor chunks Status: 18 unchanged Other — 16.3 MB (baseline 16.3 MB) • 🔴 +49 BBundles that do not match a named category
Status: 70 added / 70 removed / 224 unchanged ⚡ Performance Report
Show regressions
All metrics
Historical variance (last 15 runs)
Trend (last 15 commits on main)
Raw data{
"timestamp": "2026-09-01T06:10:54.283Z",
"gitSha": "4a1290dc7a0582b8787530263355a2b31932f4ce",
"branch": "glary/drop-duplicate-assertion-rum-events",
"measurements": [
{
"name": "canvas-idle",
"durationMs": 2034.6059999999397,
"styleRecalcs": 10,
"styleRecalcDurationMs": 6.086000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 510.81,
"heapDeltaBytes": -8265068,
"heapUsedBytes": 60539304,
"domNodes": -279,
"jsHeapTotalBytes": 4448256,
"scriptDurationMs": 8.168999999999999,
"eventListeners": -155,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-idle",
"durationMs": 2014.2209999999068,
"styleRecalcs": 10,
"styleRecalcDurationMs": 5.802,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 418.91299999999995,
"heapDeltaBytes": -10020968,
"heapUsedBytes": 59241476,
"domNodes": -278,
"jsHeapTotalBytes": 5758976,
"scriptDurationMs": 7.262000000000001,
"eventListeners": -155,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1702.8189999999768,
"styleRecalcs": 75,
"styleRecalcDurationMs": 25.7,
"layouts": 14,
"layoutDurationMs": 3.2790000000000004,
"taskDurationMs": 668.127,
"heapDeltaBytes": -7661828,
"heapUsedBytes": 62442664,
"domNodes": -277,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 103.84599999999999,
"eventListeners": -155,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1734.1209999999592,
"styleRecalcs": 73,
"styleRecalcDurationMs": 24.104,
"layouts": 14,
"layoutDurationMs": 3.293,
"taskDurationMs": 688.1089999999999,
"heapDeltaBytes": 4630008,
"heapUsedBytes": 74613984,
"domNodes": -283,
"jsHeapTotalBytes": 6021120,
"scriptDurationMs": 107.631,
"eventListeners": -185,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1690.1150000001053,
"styleRecalcs": 30,
"styleRecalcDurationMs": 12.893,
"layouts": 6,
"layoutDurationMs": 1.167,
"taskDurationMs": 296.977,
"heapDeltaBytes": 4425496,
"heapUsedBytes": 74376464,
"domNodes": 78,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 9.814000000000002,
"eventListeners": 19,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1657.4780000000828,
"styleRecalcs": 31,
"styleRecalcDurationMs": 12.812000000000001,
"layouts": 6,
"layoutDurationMs": 1.149,
"taskDurationMs": 303.531,
"heapDeltaBytes": 4469828,
"heapUsedBytes": 74303068,
"domNodes": 77,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 9.42,
"eventListeners": 19,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-clipping",
"durationMs": 515.8769999999322,
"styleRecalcs": 11,
"styleRecalcDurationMs": 5.171999999999998,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 356.39899999999994,
"heapDeltaBytes": 17857604,
"heapUsedBytes": 87937704,
"domNodes": 18,
"jsHeapTotalBytes": 4980736,
"scriptDurationMs": 119.61400000000002,
"eventListeners": 2,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-clipping",
"durationMs": 543.6110000000554,
"styleRecalcs": 12,
"styleRecalcDurationMs": 6.100000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 356.644,
"heapDeltaBytes": 18046932,
"heapUsedBytes": 88027356,
"domNodes": 20,
"jsHeapTotalBytes": 5505024,
"scriptDurationMs": 125.301,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-idle",
"durationMs": 2035.8290000001489,
"styleRecalcs": 6,
"styleRecalcDurationMs": 3.5250000000000004,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 491.6399999999999,
"heapDeltaBytes": -13003476,
"heapUsedBytes": 87110920,
"domNodes": -268,
"jsHeapTotalBytes": -1617920,
"scriptDurationMs": 31.091999999999995,
"eventListeners": -151,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-idle",
"durationMs": 2027.8559999999288,
"styleRecalcs": 6,
"styleRecalcDurationMs": 4.249000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 525.2259999999999,
"heapDeltaBytes": 11440412,
"heapUsedBytes": 86349288,
"domNodes": -269,
"jsHeapTotalBytes": -1654784,
"scriptDurationMs": 29.358,
"eventListeners": -183,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-pan",
"durationMs": 2149.6660000000247,
"styleRecalcs": 67,
"styleRecalcDurationMs": 10.027000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1280.302,
"heapDeltaBytes": 13002460,
"heapUsedBytes": 88123912,
"domNodes": -277,
"jsHeapTotalBytes": 667648,
"scriptDurationMs": 601.0600000000001,
"eventListeners": -181,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "large-graph-pan",
"durationMs": 2153.498000000127,
"styleRecalcs": 67,
"styleRecalcDurationMs": 10.374,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1300.398,
"heapDeltaBytes": 11806412,
"heapUsedBytes": 88456612,
"domNodes": -271,
"jsHeapTotalBytes": -380928,
"scriptDurationMs": 600.8209999999999,
"eventListeners": -181,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-zoom",
"durationMs": 3258.092999999917,
"styleRecalcs": 62,
"styleRecalcDurationMs": 11.268,
"layouts": 60,
"layoutDurationMs": 12.149000000000001,
"taskDurationMs": 1379.3739999999998,
"heapDeltaBytes": 8046692,
"heapUsedBytes": 75336516,
"domNodes": 4,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 605.233,
"eventListeners": 8,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-zoom",
"durationMs": 3136.7640000000847,
"styleRecalcs": 62,
"styleRecalcDurationMs": 11.037999999999998,
"layouts": 60,
"layoutDurationMs": 12.267,
"taskDurationMs": 1352.6470000000002,
"heapDeltaBytes": 11751572,
"heapUsedBytes": 79519444,
"domNodes": 4,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 581.0490000000001,
"eventListeners": 8,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "legacy-node-drag",
"durationMs": 2320.9999999999127,
"styleRecalcs": 42,
"styleRecalcDurationMs": 5.229000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1445.189,
"heapDeltaBytes": 13599528,
"heapUsedBytes": 80770828,
"domNodes": 4,
"jsHeapTotalBytes": 1204224,
"scriptDurationMs": 691.152,
"eventListeners": 186,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "legacy-node-drag",
"durationMs": 2264.9229999999534,
"styleRecalcs": 41,
"styleRecalcDurationMs": 4.997999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1506.748,
"heapDeltaBytes": 17717788,
"heapUsedBytes": 85126452,
"domNodes": -278,
"jsHeapTotalBytes": 417792,
"scriptDurationMs": 705.4390000000001,
"eventListeners": 184,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "minimap-idle",
"durationMs": 2015.041999999994,
"styleRecalcs": 5,
"styleRecalcDurationMs": 4.384000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 488.129,
"heapDeltaBytes": -14511808,
"heapUsedBytes": 87270196,
"domNodes": -270,
"jsHeapTotalBytes": -1617920,
"scriptDurationMs": 28.121999999999993,
"eventListeners": -151,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "minimap-idle",
"durationMs": 2003.9809999998397,
"styleRecalcs": 5,
"styleRecalcDurationMs": 2.9430000000000014,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 460.907,
"heapDeltaBytes": 22086156,
"heapUsedBytes": 88224956,
"domNodes": 10,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 29.319,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 529.6910000001844,
"styleRecalcs": 47,
"styleRecalcDurationMs": 7.5310000000000015,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 314.327,
"heapDeltaBytes": 13680636,
"heapUsedBytes": 85754164,
"domNodes": 20,
"jsHeapTotalBytes": 5767168,
"scriptDurationMs": 102.745,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66666666666665,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 531.011000000035,
"styleRecalcs": 47,
"styleRecalcDurationMs": 7.598999999999998,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 320.315,
"heapDeltaBytes": 14745116,
"heapUsedBytes": 87063240,
"domNodes": 20,
"jsHeapTotalBytes": 5767168,
"scriptDurationMs": 102.833,
"eventListeners": 8,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-idle",
"durationMs": 2018.3009999999513,
"styleRecalcs": 8,
"styleRecalcDurationMs": 4.986999999999998,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 413.526,
"heapDeltaBytes": 7409028,
"heapUsedBytes": 72444476,
"domNodes": -298,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 7.369000000000001,
"eventListeners": -233,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-idle",
"durationMs": 1998.119000000088,
"styleRecalcs": 11,
"styleRecalcDurationMs": 6.827,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 424.18100000000004,
"heapDeltaBytes": -5357132,
"heapUsedBytes": 66548532,
"domNodes": -293,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 7.385999999999998,
"eventListeners": -201,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1685.1850000000468,
"styleRecalcs": 76,
"styleRecalcDurationMs": 25.716999999999995,
"layouts": 18,
"layoutDurationMs": 3.9189999999999996,
"taskDurationMs": 659.3679999999999,
"heapDeltaBytes": -12059500,
"heapUsedBytes": 59869168,
"domNodes": -297,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 75.42,
"eventListeners": -201,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.670000000000012,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1693.4929999999895,
"styleRecalcs": 77,
"styleRecalcDurationMs": 27.651000000000003,
"layouts": 18,
"layoutDurationMs": 4.107,
"taskDurationMs": 691.4119999999999,
"heapDeltaBytes": -10175876,
"heapUsedBytes": 61741192,
"domNodes": -296,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 78.75999999999999,
"eventListeners": -201,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-transition-enter",
"durationMs": 924.3180000000848,
"styleRecalcs": 19,
"styleRecalcDurationMs": 24.298999999999996,
"layouts": 15,
"layoutDurationMs": 10.606000000000002,
"taskDurationMs": 725.9830000000001,
"heapDeltaBytes": -7462728,
"heapUsedBytes": 85924764,
"domNodes": 13751,
"jsHeapTotalBytes": 11010048,
"scriptDurationMs": 23.015000000000008,
"eventListeners": 2381,
"totalBlockingTimeMs": 117,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "viewport-pan-sweep",
"durationMs": 8201.458999999886,
"styleRecalcs": 250,
"styleRecalcDurationMs": 30.65,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 4432.323,
"heapDeltaBytes": 13765896,
"heapUsedBytes": 89062880,
"domNodes": -265,
"jsHeapTotalBytes": -380928,
"scriptDurationMs": 1936.356,
"eventListeners": -167,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "viewport-pan-sweep",
"durationMs": 8233.76200000007,
"styleRecalcs": 250,
"styleRecalcDurationMs": 32.519,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 4538.119,
"heapDeltaBytes": -6783548,
"heapUsedBytes": 88985592,
"domNodes": -263,
"jsHeapTotalBytes": -868352,
"scriptDurationMs": 1990.3429999999998,
"eventListeners": -167,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.80000000000109
},
{
"name": "vue-large-graph-idle",
"durationMs": 12691.931000000068,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 12047.787999999999,
"heapDeltaBytes": -26615280,
"heapUsedBytes": 174095008,
"domNodes": -6548,
"jsHeapTotalBytes": -5545984,
"scriptDurationMs": 221.918,
"eventListeners": -7475,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-idle",
"durationMs": 12677.098999999998,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 12057.646999999999,
"heapDeltaBytes": -35944348,
"heapUsedBytes": 173924680,
"domNodes": -6548,
"jsHeapTotalBytes": -528384,
"scriptDurationMs": 201.111,
"eventListeners": -7473,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.223333333333237,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-pan",
"durationMs": 15697.34399999993,
"styleRecalcs": 84,
"styleRecalcDurationMs": 14.133000000000006,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 15159.880000000001,
"heapDeltaBytes": -36206472,
"heapUsedBytes": 165799888,
"domNodes": -6548,
"jsHeapTotalBytes": 3141632,
"scriptDurationMs": 770.8530000000001,
"eventListeners": -7469,
"totalBlockingTimeMs": 24,
"frameDurationMs": 17.220000000000073,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-pan",
"durationMs": 15838.591000000179,
"styleRecalcs": 86,
"styleRecalcDurationMs": 14.097000000000026,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 15240.827,
"heapDeltaBytes": -16254324,
"heapUsedBytes": 186401544,
"domNodes": -6548,
"jsHeapTotalBytes": 737280,
"scriptDurationMs": 761.875,
"eventListeners": -7471,
"totalBlockingTimeMs": 9,
"frameDurationMs": 17.223333333333358,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-node-resize-workload",
"durationMs": 4569.467999999915,
"styleRecalcs": 20,
"styleRecalcDurationMs": 60.84999999999999,
"layouts": 20,
"layoutDurationMs": 21.098999999999993,
"taskDurationMs": 4425.492000000001,
"heapDeltaBytes": -40249356,
"heapUsedBytes": 172893080,
"domNodes": -6548,
"jsHeapTotalBytes": -6217728,
"scriptDurationMs": 76.68900000000001,
"eventListeners": -7471,
"totalBlockingTimeMs": 41,
"frameDurationMs": 17.216666666666665,
"p95FrameDurationMs": 16.80000000000109
},
{
"name": "vue-node-resize-workload",
"durationMs": 4685.660999999982,
"styleRecalcs": 20,
"styleRecalcDurationMs": 60.663000000000025,
"layouts": 20,
"layoutDurationMs": 21.898,
"taskDurationMs": 4562.641,
"heapDeltaBytes": -28147936,
"heapUsedBytes": 174410884,
"domNodes": -6548,
"jsHeapTotalBytes": 11264000,
"scriptDurationMs": 98.96500000000002,
"eventListeners": -7471,
"totalBlockingTimeMs": 84,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "workflow-execution",
"durationMs": 447.257000000036,
"styleRecalcs": 19,
"styleRecalcDurationMs": 16.973,
"layouts": 4,
"layoutDurationMs": 1.532,
"taskDurationMs": 95.143,
"heapDeltaBytes": 5180000,
"heapUsedBytes": 72287432,
"domNodes": 151,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 6.767,
"eventListeners": 97,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666682,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "workflow-execution",
"durationMs": 461.58000000014,
"styleRecalcs": 17,
"styleRecalcDurationMs": 17.773999999999997,
"layouts": 3,
"layoutDurationMs": 1.7810000000000001,
"taskDurationMs": 96.11999999999999,
"heapDeltaBytes": 5174104,
"heapUsedBytes": 73597480,
"domNodes": 144,
"jsHeapTotalBytes": 786432,
"scriptDurationMs": 6.790000000000001,
"eventListeners": 97,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.663333333333338,
"p95FrameDurationMs": 16.700000000000273
}
]
} |
|
Warning Review limit reachedNext included review available in 7 minutes. View limit detailsLimit details: You’ve used all 4 included reviews currently available. Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available. Review configuration: ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Team Run ID: 📒 Files selected for processing (4)
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Plus Run ID: 📒 Files selected for processing (9)
Included review availability: Your plan provides up to 10 included reviews per hour; 2 remain after this review. 📝 WalkthroughWalkthroughThe change standardizes assertion failure messages, exposes assertion reporter state, filters duplicate console assertions from Datadog RUM, and removes dynamic identifiers from selected diagnostics. ChangesAssertion observability
Estimated code review effort: 3 (Moderate) | ~20 minutes Merge Risk: ⚪ Minimal · up to This change removes duplicate assertion telemetry and prevents unbounded assertion details from being sent, while preserving assertion reporting through the registered reporter; no actionable merge-blocking risk remains. Sequence Diagram(s)sequenceDiagram
participant Assertion
participant AssertReporter
participant Console
participant DatadogRUM
Assertion->>AssertReporter: Report assertion failure
Assertion->>Console: Emit assertion message
Console->>DatadogRUM: Submit console error
DatadogRUM->>AssertReporter: Check reporter state
DatadogRUM-->>Console: Drop duplicate assertion error
Important Pre-merge checks failedPlease resolve all errors before merging. Addressing warnings is optional. ❌ Failed checks (1 inconclusive)
✅ Passed checks (6 passed)
Full details: End-To-End Regression Coverage For FixesExplanation The changed-file list is available, but the PR title and commit subjects are not available in the review context. The custom check requires an inconclusive result when commit subjects are unavailable, because the required bug-fix signal cannot be verified without guessing. Full details: Adr Compliance For Entity/Litegraph ChangesExplanation PASS. The changed-file list includes Full details: Description checkExplanation The description clearly explains the duplicate-reporting problem, implementation changes, verification, and follow-ups. It does not use every template heading or the requested What, Breaking, and Dependencies fields, but it provides the required information in equivalent sections. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
|
CodeRabbit approved with no actionable comments, so there is nothing to change. Supplying the context the one inconclusive pre-merge check said was missing from its review context. Title: Commits:
Why there is no The rest of the diff is assertion message strings, which have no user-visible surface; the only thing that renders one is the nightly toast, and it renders Coverage sits at the level that can actually observe the behavior:
TL;DR: the direct filter cases are covered, but the permanent suite does not exercise the ordering contract that makes “exactly once” true: real Full context for agent readerssuggestion: would it be beneficial to commit the integration regression described above into This is the remaining evidence gap I found; the static-message and schema fail-closed changes themselves preserve the documented invariant behavior. Glossary: RUM = Datadog Real User Monitoring; reporter = the registered production assertion sink; console copy = Datadog’s automatic capture of |
Codecov Report❌ Patch coverage is
@@ Coverage Diff @@
## main #16333 +/- ##
==========================================
+ Coverage 81.25% 83.26% +2.00%
==========================================
Files 2402 2023 -379
Lines 123442 116972 -6470
Branches 34492 32452 -2040
==========================================
- Hits 100300 97394 -2906
+ Misses 22543 19222 -3321
+ Partials 599 356 -243
Flags with carried forward coverage won't be shown. Click here to find out more.
... and 429 files with indirect coverage changes 🚀 New features to boost your workflow:
|
Full context for agent readersThe assertion/RUM dedupe keeps a typed thrown error while removing dynamic values from the central invariant report. CI is green at head 1e34bdc; draft status was preserved. The review checked the PR diff against |
|
Replying to the No code changes — reading it as a status attestation (mergeability, CI, reviewer assignment) rather than a change request. The summary is accurate: Since the lens is adversarial, the sharpest question this diff invites is whether the new drop rule can silence the only copy of an assertion. It can't, and the reason is worth stating explicitly because
So in every build where the drop rule can fire, the reporter is already forwarding the assertion through One bounded edge I deliberately left alone: Verified locally at |
dante01yoon
left a comment
There was a problem hiding this comment.
Draft review. I independently verified the premise and it holds, and the PR body's self-disclosure is unusually complete — both of the residual holes I went looking for are already named in "Follow-ups, deliberately not in scope". Two things left, neither of which I could settle from the code alone.
Premise verified
assert() calls console.error(formatted) before reporter?.(formatted), RUM collects console errors as error.source === 'console', and reportError produces the custom-source copy — so the double-report is real. The ordering also makes the new guard work: the console echo hits beforeSend synchronously, at which point the boot-time reporter is already installed, so it is dropped and only the tagged copy survives.
I also checked the two things that would have made this wrong and did not:
hasAssertReporter()reads module state in@/base/assert, and the newdatadogRumBeforeSend→assertimport edge could have created a second module instance. You verified this in a live runtime; I note the failure mode is fail-open (double reporting, i.e. today's behaviour) rather than silent loss, so the downside is bounded either way.schemaGuard.tsstill builds the interpolatedmessageand passes it toFollowerSchemaError, which I thought might leak the peer-influencedString(found)to RUM through the exception path instead. It does not:layoutFollowerBridge.ts:198-209catchesFollowerSchemaErrorspecifically, stores it, and re-dispatchesfoundas a structuredschema_errordetail. It never propagates uncaught.messageis not dead either.
Agreeing with the message-stripping half
Worth saying explicitly since it reads as scope creep at first glance: assert()'s docstring on main already required it — "must be a static description of the invariant. Never interpolate user data (workflow names, paths, prompts) into it." changeTracker.ts shipping a user's workflowPath to Datadog, Sentry and a nightly toast was a straight violation of that, and keeping the path in the local dedup key while removing it from the message is the right split.
Two inline items. Nothing blocking from me on the draft; the question: is the one I would want answered before this leaves draft.
Reviewed at 1e34bdc7a6e5da4267bff5b6609c103ae6e6ee85. No tests run locally — the diff is small enough to read, and your mutation-testing notes cover the guards I would have poked at.
|
@coderabbitai review |
|
|
Two housekeeping items from this pass, neither of which needs a code change. The PR body is stale on the exact point @dante01yoon raised.
CodeRabbit has not reviewed the current head. The Head is unchanged at |
DrJKL
left a comment
There was a problem hiding this comment.
The earlier RUM-capability concern is addressed: console assertion events are suppressed only when the installed reporter explicitly forwards to RUM, while non-RUM and absent reporters keep the console event. Raw telemetry sinks remain confined to telemetry internals. Focused tests pass 89/89 and the regression is red on base; typecheck and scoped lint/format pass. The deferred recurrence-count concern remains non-blocking.
|
Thanks @DrJKL. No code changes in response to this review — flagging that explicitly so it isn't mistaken for an oversight. The approval confirms the RUM-capability fix and marks the recurrence-count item non-blocking; neither asks for a change, and the branch is unchanged at the head you approved ( What I did do is re-verify the approval's claims first-hand rather than inherit them. Gates at
|
|
Follow-up queued: assertion telemetry fidelity. Full context for agent readersThe program queue now carries the frontend #16333 follow-up task, The audit found no open or merged PR whose title or body back-references #16333. The carrier must link back here when opened. |
…6333) *PR Created by the Glary-Bot Agent* --- On Cloud, a failed `assert()` reaches Datadog twice. RUM collects `console.error` on its own, and `assert()` calls it before handing the failure to the reporter that #15189 wired up — so every assertion arrives once untagged from the console and once tagged via `reportError`. The untagged copy also bypasses `reportAssertFailure`'s dedup and its 20-per-session cap, so a render-loop invariant floods RUM regardless of either safeguard. ## Changes - `datadogRumBeforeSend` drops the console-sourced copy of an assertion, but only while a reporter is registered to produce the tagged one, so a boot-path assertion is never silenced entirely. `assert.ts` exports `ASSERTION_FAILURE_PREFIX` and `hasAssertReporter()` for that check; `assert()`'s own output is byte-identical. - Four assertion messages stopped interpolating unbounded values. `reportAssertFailure` dedups on exact message and caps a session at 20 distinct ones, so an interpolated value made each occurrence its own message. `unregisterNodeState` is the sharp case: `detachAllNodesFromStores` loops every node in a graph, so clearing a 20-node graph under identity drift burned the entire session budget in one call — and once the cap is spent no tagged report is produced, while the console echo is now dropped, leaving nothing at all. Static messages hold cardinality at 6 across all 5 call sites. - `nodeShellState.ts` x2 (`node.id`, `rootGraphId`), `inputSpecUtil.ts` (`spec.name`), `schemaGuard.ts` (`String(found)`, read `unknown` from a merged Yjs doc and therefore peer-influenced). `FollowerSchemaError` still carries `found` as a structured field. - `changeTracker.ts` stops interpolating `workflowPath`, which was shipping a user's workflow path to Datadog, Sentry, and the nightly toast. It stays in the local dedup key. All four were already required by `assert()`'s own docstring on `main` — "must be a static description of the invariant. Never interpolate user data (workflow names, paths, prompts) into it." ## Verification - `pnpm typecheck`, `pnpm lint` (0 errors), `pnpm format:check`, `pnpm knip` clean. 100 tests across the 7 affected files, plus the full CRDT suite at 160. - Mutation-tested each new guard: reverting the RUM filter fails `drops the console echo...`; restoring either interpolated `changeTracker` message fails its pinned assertion; restoring `schemaGuard`'s interpolated message fails both the positive and the new negative matcher. - Integration run through real `assert` -> real `reportAssertFailure` -> real `reportError`, capturing only the sinks: one assertion produces exactly one RUM-bound event; 100 repeats still one; 50 distinct node detachments produce one and leave later unrelated assertions deliverable. Under the old interpolated form the same 50 exhaust the cap and a subsequent unrelated assertion produces zero events. - Real browser against the dev server: importing `/src/base/assert.ts` returns `hasAssertReporter() === true`, so the app's own registration is visible to the same module instance, ruling out an import cycle from the new `datadogRumBeforeSend` -> `assert` edge. In that live runtime `rumBeforeSend` returns false for a console-sourced assertion, true for a custom-sourced one, and true for an ordinary console error. ## Follow-ups, deliberately not in scope - `schemaGuard`'s `found` and `inputSpecUtil`'s `specName` no longer reach telemetry at all. The right home is `reportError`'s existing structured `context`/`tags`, which are indexed and do not participate in message grouping — either called directly or by widening `AssertReporter` with an optional context. It is a design change to `assert()` rather than a fix to this bug. That follow-up should also preserve coarse recurrence information (for example thresholded counts) so a one-off and a render-loop storm remain distinguishable without restoring per-occurrence cardinality. - `reportAssertFailure` adds to `reportedMessages` before `reportError` is called, and `reportError` returns early once any sink is live — so a report raised while Sentry is up but RUM is still initializing burns its dedup slot without reaching RUM, and is never buffered for the later flush. Predates this branch, in files it does not touch. ## Context Supersedes #15671, which targeted the same problem before #15189 landed the reporting path on `main` and is now largely redundant against it. #15671 should be closed. --------- Co-authored-by: Glary-Bot <glary-bot@users.noreply.github.com> Co-authored-by: Pipeline Agent <agent@pipeline.local>
|
Replying to the "Follow-up queued: assertion telemetry fidelity" audit comment (5510464241) — posting here because that comment is file-level and has no diff-line parent to thread under. Acknowledged. No code change from me: #16333 is merged and
Two things the carrier should absorb that the audit summary doesn't name: 1. 2. Sequencing hazard: don't close #15671 before the carrier lands. This PR's description says "#15671 should be closed", and it is still open — but "largely redundant" undersells it. One caveat on that lift: #15671 routes Understood on the back-reference — the carrier PR should link to #16333 when opened. I haven't opened it, since it's @DrJKL's. |
|
Follow-up carrier updated: #15671. Structured assertion context and reliable cloud RUM delivery are now implemented on current Full context for agent readersThe existing carrier was rebased through a merge with current Carrier head: |
|
Replying to the "Follow-up carrier updated: #15671" comment (5511021603) — posting here because that comment is file-level and has no diff-line parent to thread under. Checked out All four claims hold:
Verification reproduced locally: 52 tests green across the 7 affected files, Also worth saying the narrowing was the right call — dropping #15671's One regression I'd close before merge.
mockIsCloud.value = true
sentryLive(false); datadogLive(false)
reportError(new Error('cold boot'), { errorType: 'invariant_assert' })
sentryLive(true); datadogLive(true)
addError.mockImplementationOnce(() => { throw new Error('rum exploded') })
flushErrorReports()
flushErrorReports()
expect(captureException).toHaveBeenCalledOnce() // received 2This is new here: Related, and pre-existing — flagging because one fix covers both. Both fall out of letting let sentry = false
if (sentryLive) {
try { captureException(/* … */); sentry = true }
catch (e) { console.error('[reportError] sentry sink failed', e) }
}
let datadog = false
if (datadogLive) {
try { datadogRum.addError(/* … */); datadog = true }
catch (e) { console.error('[reportError] datadog sink failed', e) }
}
return { sentry, datadog }The re-queue then always carries accurate delivery state, and one sink failing can't suppress the other. I haven't pushed any of this — the carrier is @DrJKL's lane and under active work, so it isn't mine to commit to. Happy to hand over the probe as a proper regression test if useful. |
Preserve successful sink delivery state and continue dispatching when another telemetry sink throws. Source: #16333 (comment)
## Summary Preserve actionable diagnostics for assertion failures across Sentry and Datadog, completing the review follow-up from #16333 without changing assertion message grouping. ## Changes - **Structured diagnostics**: `assert()` accepts optional context and forwards `found`, `specName`, node ID, and root graph ID from the reviewed invariant sites. - **Recurrence signal**: repeated assertions emit at coarse counts 1, 10, 100, and 1000 while retaining the 20-distinct-message bound. - **Cloud delivery**: when Sentry initializes before Datadog RUM, `reportError` still queues the RUM leg without sending a duplicate Sentry event. - **Privacy boundary**: this does not automatically restore `workflowPath`; that remains deliberately excluded pending a separate privacy decision. ## Review Focus Please review sink-specific delivery state in `reportError` and the structured context emitted by graph/schema assertions. Follow-up to #16333. ## Evidence - `pnpm vitest run src/base/assert.test.ts src/platform/telemetry/assertFailureReporter.test.ts src/platform/telemetry/reportError.test.ts src/schemas/nodeDef/inputSpecUtil.test.ts src/workbench/extensions/agent/crdt/schemaGuard.test.ts src/core/graph/nodeShell/nodeShellState.test.ts` — 6 files, 42 tests passed. - `pnpm vitest run src/base/assert.test.ts src/platform/telemetry/assertFailureReporter.test.ts src/platform/telemetry/reportError.test.ts` — repeat run: 3 files, 22 tests passed. - `pnpm typecheck` — passed. - Targeted ESLint and oxfmt checks over all 13 changed files — passed. - Pre-commit hooks: lint-staged formatting, full lint, typecheck, and knip passed. - `git diff --check origin/main` — passed. <!-- authored-by:agent lane-act-fe-16333-followup --> --------- Co-authored-by: Glary Bot <bot@comfy.org> Co-authored-by: GitHub Action <action@github.com> Co-authored-by: Amp <amp@ampcode.com>
#17425) Backport of the telemetry assertion-fidelity fix to core/1.53, with its prerequisites. - Backports [preserve assertion diagnostic fidelity](#15671) (merged to main as `3d7c5d4df3`). - Pulls in the three commits it depends on that were not yet on core/1.53: [report each assertion failure to RUM exactly once](#16333), [report stable native Datadog error types](#17085), and [report layout listener failures to telemetry](#17040). - Targeted vitest (40 files, 468 tests) and `pnpm typecheck` pass on this branch. <details><summary>Full context for agent readers</summary> Cherry-picked with `-x`, in dependency order, onto `0c0a1ec66a` (origin/core/1.53): 1. `83fb266a6e` (RUM exactly-once) — required by the assertion code path. 2. `74564e32d3` (stable native error types) — required by the assertion code path. 3. `3d7c5d4df3` (the fix itself). 4. `cf15ffd685` (layout listener failures) — pulled in because three tests in `reportError.test.ts` introduced by the fix rely on the `reportError()` layout-listener behavior from that change; without it the backport is red. Not pulled: `606fb5eedf` (already present on 1.53 via the earlier main.ts backport), and unrelated main.ts changes `ee86e426ef` / `05672356c6`. A matching cloud/1.53 backport follows separately. <!-- authored-by:agent supreme-operator w3-w42 --> </details> --------- Co-authored-by: Glary-Bot <glary-bot@users.noreply.github.com> Co-authored-by: Pipeline Agent <agent@pipeline.local> Co-authored-by: ShihChi Huang <huge.huang@gmail.com> Co-authored-by: huang47 <157390+huang47@users.noreply.github.com> Co-authored-by: Glary Bot <bot@comfy.org> Co-authored-by: GitHub Action <action@github.com> Co-authored-by: Amp <amp@ampcode.com> Co-authored-by: Alexander Brown <drjkl@comfy.org>
…ckport) (#17427) Backport of the telemetry assertion-fidelity fix to cloud/1.53, with its prerequisites. Sibling of the already-merged [core/1.53 backport](#17425). - Backports [preserve assertion diagnostic fidelity](#15671) (merged to main as `3d7c5d4df3`). - Pulls in the three commits it depends on that were not yet on cloud/1.53: [report each assertion failure to RUM exactly once](#16333), [report stable native Datadog error types](#17085), and [report layout listener failures to telemetry](#17040). - Targeted vitest (`src/platform/telemetry`: 29 files, 393 tests) passes on this branch. <details><summary>Full context for agent readers</summary> Cherry-picked with `-x`, in dependency order, onto `35c46f0ec9` (origin/cloud/1.53): 1. `83fb266a6e` (RUM exactly-once) — required by the assertion code path. 2. `74564e32d3` (stable native error types) — required by the assertion code path. 3. `3d7c5d4df3` (the fix itself). 4. `cf15ffd685` (layout listener failures) — required because three `reportError.test.ts` tests introduced by the fix rely on the layout-listener `reportError()` behavior. 5. `e2a2cd419f` alignment commit, cloud-only: the cherry-pick of `cf15ffd685` conflicted in `assertFailureReporter.ts` and the automatic resolution dropped `occurrenceCount` from the reported `context`. This commit restores `context: { ...context, occurrenceCount }` and the matching test expectation, so the file now matches the core/1.53 version byte-for-byte. Not pulled: `606fb5eedf` (already present on 1.53), and unrelated main.ts changes `ee86e426ef` / `05672356c6`. <!-- authored-by:agent supreme-operator w3-w42 --> </details> --------- Co-authored-by: Glary-Bot <glary-bot@users.noreply.github.com> Co-authored-by: Pipeline Agent <agent@pipeline.local> Co-authored-by: ShihChi Huang <huge.huang@gmail.com> Co-authored-by: huang47 <157390+huang47@users.noreply.github.com> Co-authored-by: Glary Bot <bot@comfy.org> Co-authored-by: GitHub Action <action@github.com> Co-authored-by: Amp <amp@ampcode.com> Co-authored-by: Alexander Brown <drjkl@comfy.org>
Preserve actionable diagnostics for assertion failures across Sentry and Datadog, completing the review follow-up from #16333 without changing assertion message grouping. - **Structured diagnostics**: `assert()` accepts optional context and forwards `found`, `specName`, node ID, and root graph ID from the reviewed invariant sites. - **Recurrence signal**: repeated assertions emit at coarse counts 1, 10, 100, and 1000 while retaining the 20-distinct-message bound. - **Cloud delivery**: when Sentry initializes before Datadog RUM, `reportError` still queues the RUM leg without sending a duplicate Sentry event. - **Privacy boundary**: this does not automatically restore `workflowPath`; that remains deliberately excluded pending a separate privacy decision. Please review sink-specific delivery state in `reportError` and the structured context emitted by graph/schema assertions. Follow-up to #16333. - `pnpm vitest run src/base/assert.test.ts src/platform/telemetry/assertFailureReporter.test.ts src/platform/telemetry/reportError.test.ts src/schemas/nodeDef/inputSpecUtil.test.ts src/workbench/extensions/agent/crdt/schemaGuard.test.ts src/core/graph/nodeShell/nodeShellState.test.ts` — 6 files, 42 tests passed. - `pnpm vitest run src/base/assert.test.ts src/platform/telemetry/assertFailureReporter.test.ts src/platform/telemetry/reportError.test.ts` — repeat run: 3 files, 22 tests passed. - `pnpm typecheck` — passed. - Targeted ESLint and oxfmt checks over all 13 changed files — passed. - Pre-commit hooks: lint-staged formatting, full lint, typecheck, and knip passed. - `git diff --check origin/main` — passed. <!-- authored-by:agent lane-act-fe-16333-followup --> --------- Co-authored-by: Glary Bot <bot@comfy.org> Co-authored-by: GitHub Action <action@github.com> Co-authored-by: Amp <amp@ampcode.com> (cherry picked from commit 0fd0b75)
Preserve actionable diagnostics for assertion failures across Sentry and Datadog, completing the review follow-up from #16333 without changing assertion message grouping. - **Structured diagnostics**: `assert()` accepts optional context and forwards `found`, `specName`, node ID, and root graph ID from the reviewed invariant sites. - **Recurrence signal**: repeated assertions emit at coarse counts 1, 10, 100, and 1000 while retaining the 20-distinct-message bound. - **Cloud delivery**: when Sentry initializes before Datadog RUM, `reportError` still queues the RUM leg without sending a duplicate Sentry event. - **Privacy boundary**: this does not automatically restore `workflowPath`; that remains deliberately excluded pending a separate privacy decision. Please review sink-specific delivery state in `reportError` and the structured context emitted by graph/schema assertions. Follow-up to #16333. - `pnpm vitest run src/base/assert.test.ts src/platform/telemetry/assertFailureReporter.test.ts src/platform/telemetry/reportError.test.ts src/schemas/nodeDef/inputSpecUtil.test.ts src/workbench/extensions/agent/crdt/schemaGuard.test.ts src/core/graph/nodeShell/nodeShellState.test.ts` — 6 files, 42 tests passed. - `pnpm vitest run src/base/assert.test.ts src/platform/telemetry/assertFailureReporter.test.ts src/platform/telemetry/reportError.test.ts` — repeat run: 3 files, 22 tests passed. - `pnpm typecheck` — passed. - Targeted ESLint and oxfmt checks over all 13 changed files — passed. - Pre-commit hooks: lint-staged formatting, full lint, typecheck, and knip passed. - `git diff --check origin/main` — passed. <!-- authored-by:agent lane-act-fe-16333-followup --> --------- Co-authored-by: Glary Bot <bot@comfy.org> Co-authored-by: GitHub Action <action@github.com> Co-authored-by: Amp <amp@ampcode.com>
PR Created by the Glary-Bot Agent
On Cloud, a failed
assert()reaches Datadog twice. RUM collectsconsole.erroron its own, andassert()calls it before handing the failure to the reporter that #15189 wired up — so every assertion arrives once untagged from the console and once tagged viareportError. The untagged copy also bypassesreportAssertFailure's dedup and its 20-per-session cap, so a render-loop invariant floods RUM regardless of either safeguard.Changes
datadogRumBeforeSenddrops the console-sourced copy of an assertion, but only while a reporter is registered to produce the tagged one, so a boot-path assertion is never silenced entirely.assert.tsexportsASSERTION_FAILURE_PREFIXandhasAssertReporter()for that check;assert()'s own output is byte-identical.reportAssertFailurededups on exact message and caps a session at 20 distinct ones, so an interpolated value made each occurrence its own message.unregisterNodeStateis the sharp case:detachAllNodesFromStoresloops every node in a graph, so clearing a 20-node graph under identity drift burned the entire session budget in one call — and once the cap is spent no tagged report is produced, while the console echo is now dropped, leaving nothing at all. Static messages hold cardinality at 6 across all 5 call sites.nodeShellState.tsx2 (node.id,rootGraphId),inputSpecUtil.ts(spec.name),schemaGuard.ts(String(found), readunknownfrom a merged Yjs doc and therefore peer-influenced).FollowerSchemaErrorstill carriesfoundas a structured field.changeTracker.tsstops interpolatingworkflowPath, which was shipping a user's workflow path to Datadog, Sentry, and the nightly toast. It stays in the local dedup key.All four were already required by
assert()'s own docstring onmain— "must be a static description of the invariant. Never interpolate user data (workflow names, paths, prompts) into it."Verification
pnpm typecheck,pnpm lint(0 errors),pnpm format:check,pnpm knipclean. 100 tests across the 7 affected files, plus the full CRDT suite at 160.drops the console echo...; restoring either interpolatedchangeTrackermessage fails its pinned assertion; restoringschemaGuard's interpolated message fails both the positive and the new negative matcher.assert-> realreportAssertFailure-> realreportError, capturing only the sinks: one assertion produces exactly one RUM-bound event; 100 repeats still one; 50 distinct node detachments produce one and leave later unrelated assertions deliverable. Under the old interpolated form the same 50 exhaust the cap and a subsequent unrelated assertion produces zero events./src/base/assert.tsreturnshasAssertReporter() === true, so the app's own registration is visible to the same module instance, ruling out an import cycle from the newdatadogRumBeforeSend->assertedge. In that live runtimerumBeforeSendreturns false for a console-sourced assertion, true for a custom-sourced one, and true for an ordinary console error.Follow-ups, deliberately not in scope
schemaGuard'sfoundandinputSpecUtil'sspecNameno longer reach telemetry at all. The right home isreportError's existing structuredcontext/tags, which are indexed and do not participate in message grouping — either called directly or by wideningAssertReporterwith an optional context. It is a design change toassert()rather than a fix to this bug. That follow-up should also preserve coarse recurrence information (for example thresholded counts) so a one-off and a render-loop storm remain distinguishable without restoring per-occurrence cardinality.reportAssertFailureadds toreportedMessagesbeforereportErroris called, andreportErrorreturns early once any sink is live — so a report raised while Sentry is up but RUM is still initializing burns its dedup slot without reaching RUM, and is never buffered for the later flush. Predates this branch, in files it does not touch.Context
Supersedes #15671, which targeted the same problem before #15189 landed the reporting path on
mainand is now largely redundant against it. #15671 should be closed.