Repository navigation
test(agent): pin reopening an agent-generated job as a workflow - #18936
christian-byrne wants to merge 4 commits into
Conversation
…urface The agent fixtures mock /api/agent/* and the doc socket but carried no execution: /api/jobs answered an empty list forever, a submitted /api/prompt was swallowed after capture, and AgentFollowerHostSocket.send() rejects any frame that is not a doc or agent frame -- so no agent spec could assert anything that depends on a workflow actually running (a completed run with visible outputs, a failed job, a stalled job). - AgentFollowerHostSocket.sendExecution(): emits ComfyUI execution frames on the same routed /ws the doc and agent frames ride, typed against the app's own ws message shapes. - AgentExecutionHelper: stateful /api/jobs (status- and pagination-aware, via AssetsHelper), /api/prompt capture answering a fresh job id, /api/view bytes for outputs, and lifecycle drivers: startJob, completeJob (output on the node + a history row), failJob (deterministic error), stallJob (frozen running state, no terminal frame), enqueueServerRun for jobs the agent's run tool submits server-side. - MultiAutogrowRealignHarness.setUp() takes an optional settings override so rigs can pin release-channel-dependent settings (Comfy.Queue.QPOV2). - agentExecutionLifecycle.spec.ts: capability proof, three green cases, one per lifecycle. Verified locally 9/9 with --repeat-each=3; dropping the executed frame turns the completed case red.
…ws the output User-story matrix rank 26, the product's most important journey, previously with no automated coverage. Black-box on the agent harness's execution surface (#18790): the user asks the agent to run, approves the run-approval card, the run tool submits server-side (the browser never POSTs /api/prompt), and the finished output renders on the producing node and as a queue history row. Adds agentExecutionStoryFixture: the execution rig with no turn spent yet, for specs whose subject is the turn itself. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…execution surface Four stories from the user-story test matrix, black-box on the agent harness's execution surface (#18790), stacked on the rank-26 golden path: - rank 41: stopping the agent leaves its already-submitted job running - rank 42: a finished run's output shows on the producing node, not only in the assets panel - rank 99: a failed run is named to the user (queue row details) and a stalled run stays visibly running; the panel goes idle either way - rank 98 (fixture-supported half): a completed server-side generation renders a preview thumbnail from its served output; the agent-panel result-link half is deliberately deferred — it needs real reply content, not fixture-invented frames AgentExecutionHelper.enqueueServerRun gains an optional workflowId so a server-submitted job row can be attributed the way real cloud rows are. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Repository: Comfy-Org/ComfyUI_frontend/.coderabbit.yaml Review profile: ASSERTIVE Plan: Team Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Comment |
🎭 Playwright: ✅ 2128 passed, 0 failed · 2 flaky📊 Browser Reports
🌾 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 10bce7a27a53fbe30169568e61887044039aea38
Full docs: ADR-DEVEX-LINT-0015 📦 Bundle Size
⚡ Performance Report
Absolute values
Raw data{
"timestamp": "2026-09-26T01:22:18.197Z",
"gitSha": "a102bb9f970f5154774ce6c2b7584913480c2eb5",
"branch": "christian-byrne/qspec12-agent-job-reopen",
"measurements": [
{
"name": "canvas-idle",
"durationMs": 2021.2349999999333,
"styleRecalcs": 10,
"styleRecalcDurationMs": 3.71,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 445.84599999999995,
"heapDeltaBytes": -820724,
"heapUsedBytes": 61062548,
"domNodes": -228,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 8.651,
"eventListeners": -164,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "canvas-idle",
"durationMs": 2002.9930000000604,
"styleRecalcs": 11,
"styleRecalcDurationMs": 3.658,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 448.84000000000003,
"heapDeltaBytes": 6137428,
"heapUsedBytes": 68551312,
"domNodes": -228,
"jsHeapTotalBytes": 4448256,
"scriptDurationMs": 9.136999999999999,
"eventListeners": -142,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1785.5389999999716,
"styleRecalcs": 69,
"styleRecalcDurationMs": 17.465999999999998,
"layouts": 12,
"layoutDurationMs": 4.554,
"taskDurationMs": 816.769,
"heapDeltaBytes": -2442772,
"heapUsedBytes": 59945936,
"domNodes": -228,
"jsHeapTotalBytes": 5234688,
"scriptDurationMs": 156.786,
"eventListeners": -164,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "canvas-mouse-sweep",
"durationMs": 1759.1899999999896,
"styleRecalcs": 70,
"styleRecalcDurationMs": 16.653000000000002,
"layouts": 12,
"layoutDurationMs": 4.617,
"taskDurationMs": 845.671,
"heapDeltaBytes": 9272316,
"heapUsedBytes": 71661500,
"domNodes": -228,
"jsHeapTotalBytes": 5234688,
"scriptDurationMs": 152.995,
"eventListeners": -142,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1716.7409999999563,
"styleRecalcs": 31,
"styleRecalcDurationMs": 10.511000000000001,
"layouts": 6,
"layoutDurationMs": 1.2790000000000001,
"taskDurationMs": 378.293,
"heapDeltaBytes": 2208588,
"heapUsedBytes": 64551060,
"domNodes": 53,
"jsHeapTotalBytes": 4980736,
"scriptDurationMs": 14.662,
"eventListeners": 15,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "canvas-zoom-sweep",
"durationMs": 1707.6540000000477,
"styleRecalcs": 31,
"styleRecalcDurationMs": 9.901,
"layouts": 6,
"layoutDurationMs": 1.338,
"taskDurationMs": 321.5520000000001,
"heapDeltaBytes": 2157144,
"heapUsedBytes": 64578324,
"domNodes": 53,
"jsHeapTotalBytes": 4980736,
"scriptDurationMs": 11.098,
"eventListeners": 15,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "dom-widget-clipping",
"durationMs": 680.0800000000891,
"styleRecalcs": 12,
"styleRecalcDurationMs": 3.4619999999999997,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 444.66099999999994,
"heapDeltaBytes": 18265284,
"heapUsedBytes": 80747104,
"domNodes": 0,
"jsHeapTotalBytes": 4980736,
"scriptDurationMs": 166.65400000000002,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-clipping",
"durationMs": 697.10299999997,
"styleRecalcs": 14,
"styleRecalcDurationMs": 4.453,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 462.214,
"heapDeltaBytes": 19074732,
"heapUsedBytes": 86514656,
"domNodes": 0,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 169.05800000000002,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-order-dense-n200-w200-i120",
"durationMs": 1157.980000000009,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 811.375,
"heapDeltaBytes": 34790432,
"heapUsedBytes": 101425704,
"domNodes": 0,
"jsHeapTotalBytes": 19034112,
"scriptDurationMs": 63.277000000000015,
"eventListeners": 0,
"totalBlockingTimeMs": 537,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-order-dense-n200-w200-i120",
"durationMs": 1343.3979999999792,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1024.661,
"heapDeltaBytes": 14068880,
"heapUsedBytes": 100982196,
"domNodes": -259,
"jsHeapTotalBytes": 25841664,
"scriptDurationMs": 74.206,
"eventListeners": -185,
"totalBlockingTimeMs": 716,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "dom-widget-order-sparse-n1000-w20-i120",
"durationMs": 846.7039999999315,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 458.866,
"heapDeltaBytes": 13973780,
"heapUsedBytes": 90493128,
"domNodes": 0,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 138.851,
"eventListeners": 0,
"totalBlockingTimeMs": 225,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "dom-widget-order-sparse-n1000-w20-i120",
"durationMs": 844.504000000029,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 467.59099999999967,
"heapDeltaBytes": 14392564,
"heapUsedBytes": 90911296,
"domNodes": 0,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 145.602,
"eventListeners": 0,
"totalBlockingTimeMs": 226,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "extension-clean-combined-draw",
"durationMs": 8760.874999999998,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 7831.759999999999,
"heapDeltaBytes": -736264,
"heapUsedBytes": 81364592,
"domNodes": 0,
"jsHeapTotalBytes": 7340032,
"scriptDurationMs": 6490.525,
"eventListeners": 0,
"totalBlockingTimeMs": 410,
"frameDurationMs": 18.333333333333332,
"p95FrameDurationMs": 33.30000000000109
},
{
"name": "extension-clean-combined-draw",
"durationMs": 9265.510000000064,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 8342.336,
"heapDeltaBytes": 11195992,
"heapUsedBytes": 93473832,
"domNodes": 0,
"jsHeapTotalBytes": 4980736,
"scriptDurationMs": 6739.955,
"eventListeners": 0,
"totalBlockingTimeMs": 160,
"frameDurationMs": 17.77333333333336,
"p95FrameDurationMs": 33.29999999999927
},
{
"name": "extension-rgthree-prototype-draw",
"durationMs": 9347.372999999949,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 8468.483,
"heapDeltaBytes": 7576672,
"heapUsedBytes": 84877600,
"domNodes": -228,
"jsHeapTotalBytes": -8192,
"scriptDurationMs": 6834.338000000001,
"eventListeners": -164,
"totalBlockingTimeMs": 240,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 33.29999999999927
},
{
"name": "extension-rgthree-prototype-draw",
"durationMs": 8257.860999999935,
"styleRecalcs": 1,
"styleRecalcDurationMs": 0.42200000000000226,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 7592.5019999999995,
"heapDeltaBytes": -4159848,
"heapUsedBytes": 72730256,
"domNodes": -228,
"jsHeapTotalBytes": 4710400,
"scriptDurationMs": 6275.585,
"eventListeners": -164,
"totalBlockingTimeMs": 177,
"frameDurationMs": 18.889999999999965,
"p95FrameDurationMs": 33.30000000000109
},
{
"name": "large-graph-idle",
"durationMs": 2043.2389999999714,
"styleRecalcs": 8,
"styleRecalcDurationMs": 2.883,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 584.3190000000001,
"heapDeltaBytes": 2367008,
"heapUsedBytes": 83200304,
"domNodes": -228,
"jsHeapTotalBytes": -233472,
"scriptDurationMs": 37.058,
"eventListeners": -142,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "large-graph-idle",
"durationMs": 2047.0980000000054,
"styleRecalcs": 9,
"styleRecalcDurationMs": 3.6259999999999972,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 612.757,
"heapDeltaBytes": -7299984,
"heapUsedBytes": 74150508,
"domNodes": -228,
"jsHeapTotalBytes": 28672,
"scriptDurationMs": 39.744,
"eventListeners": -164,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-pan",
"durationMs": 2499.9950000000126,
"styleRecalcs": 65,
"styleRecalcDurationMs": 9.763999999999998,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1696.878,
"heapDeltaBytes": -816300,
"heapUsedBytes": 81439988,
"domNodes": -228,
"jsHeapTotalBytes": 1302528,
"scriptDurationMs": 848.371,
"eventListeners": -142,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "large-graph-pan",
"durationMs": 2251.145999999949,
"styleRecalcs": 68,
"styleRecalcDurationMs": 10.441000000000003,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1560.565,
"heapDeltaBytes": 20227060,
"heapUsedBytes": 84268928,
"domNodes": -228,
"jsHeapTotalBytes": 749568,
"scriptDurationMs": 791.38,
"eventListeners": -142,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.670000000000012,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "large-graph-zoom",
"durationMs": 3750.42499999995,
"styleRecalcs": 63,
"styleRecalcDurationMs": 12.024,
"layouts": 60,
"layoutDurationMs": 15.870000000000003,
"taskDurationMs": 1783.688,
"heapDeltaBytes": 8287952,
"heapUsedBytes": 74165668,
"domNodes": 0,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 811.105,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333332,
"p95FrameDurationMs": 16.80000000000109
},
{
"name": "large-graph-zoom",
"durationMs": 3591.9450000000097,
"styleRecalcs": 63,
"styleRecalcDurationMs": 11.425999999999998,
"layouts": 60,
"layoutDurationMs": 14.496,
"taskDurationMs": 1635.825,
"heapDeltaBytes": 8221124,
"heapUsedBytes": 74202688,
"domNodes": 0,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 771.9430000000001,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "legacy-node-drag",
"durationMs": 3203.5949999999502,
"styleRecalcs": 46,
"styleRecalcDurationMs": 8.187,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 2031.656,
"heapDeltaBytes": -712616,
"heapUsedBytes": 83787952,
"domNodes": -198,
"jsHeapTotalBytes": 409600,
"scriptDurationMs": 1008.788,
"eventListeners": 16,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "legacy-node-drag",
"durationMs": 3082.418999999959,
"styleRecalcs": 45,
"styleRecalcDurationMs": 6.490000000000003,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 1863.8000000000002,
"heapDeltaBytes": 25544820,
"heapUsedBytes": 91645336,
"domNodes": 0,
"jsHeapTotalBytes": 417792,
"scriptDurationMs": 954.6129999999999,
"eventListeners": 180,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666636,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "minimap-idle",
"durationMs": 2020.0700000000325,
"styleRecalcs": 7,
"styleRecalcDurationMs": 2.656000000000002,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 675.826,
"heapDeltaBytes": -4665356,
"heapUsedBytes": 84010400,
"domNodes": -228,
"jsHeapTotalBytes": -495616,
"scriptDurationMs": 37.926,
"eventListeners": -142,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "minimap-idle",
"durationMs": 2022.3489999999629,
"styleRecalcs": 7,
"styleRecalcDurationMs": 2.126000000000003,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 530.689,
"heapDeltaBytes": 20137104,
"heapUsedBytes": 84312344,
"domNodes": -228,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 34.164,
"eventListeners": -142,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 592.3470000000179,
"styleRecalcs": 47,
"styleRecalcDurationMs": 6.4239999999999995,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 370.845,
"heapDeltaBytes": 13133880,
"heapUsedBytes": 76647100,
"domNodes": 0,
"jsHeapTotalBytes": 5242880,
"scriptDurationMs": 146.982,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-dom-widget-clipping",
"durationMs": 582.072999999923,
"styleRecalcs": 48,
"styleRecalcDurationMs": 6.674000000000001,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 395.06700000000006,
"heapDeltaBytes": 13825076,
"heapUsedBytes": 77228036,
"domNodes": 0,
"jsHeapTotalBytes": 5505024,
"scriptDurationMs": 147.334,
"eventListeners": 6,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.800000000000182
},
{
"name": "subgraph-idle",
"durationMs": 1999.7109999999338,
"styleRecalcs": 11,
"styleRecalcDurationMs": 4.052999999999999,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 450.96999999999997,
"heapDeltaBytes": 3480516,
"heapUsedBytes": 67067012,
"domNodes": -244,
"jsHeapTotalBytes": 5234688,
"scriptDurationMs": 7.564000000000001,
"eventListeners": -188,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-idle",
"durationMs": 1990.6440000000885,
"styleRecalcs": 11,
"styleRecalcDurationMs": 3.9219999999999984,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 391.647,
"heapDeltaBytes": 25296104,
"heapUsedBytes": 87013632,
"domNodes": 0,
"jsHeapTotalBytes": 4980736,
"scriptDurationMs": 7.126,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1734.5779999999422,
"styleRecalcs": 72,
"styleRecalcDurationMs": 18.109,
"layouts": 16,
"layoutDurationMs": 5.813000000000001,
"taskDurationMs": 773.157,
"heapDeltaBytes": -5288396,
"heapUsedBytes": 58274952,
"domNodes": -244,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 105.219,
"eventListeners": -188,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.66333333333335,
"p95FrameDurationMs": 16.699999999999818
},
{
"name": "subgraph-mouse-sweep",
"durationMs": 1719.312000000059,
"styleRecalcs": 74,
"styleRecalcDurationMs": 18.72,
"layouts": 16,
"layoutDurationMs": 6.303,
"taskDurationMs": 782.845,
"heapDeltaBytes": -4472912,
"heapUsedBytes": 59172576,
"domNodes": -244,
"jsHeapTotalBytes": 5496832,
"scriptDurationMs": 107.508,
"eventListeners": -210,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "subgraph-transition-enter",
"durationMs": 1368.417999999906,
"styleRecalcs": 17,
"styleRecalcDurationMs": 27.885,
"layouts": 5,
"layoutDurationMs": 12.012000000000002,
"taskDurationMs": 877.9079999999999,
"heapDeltaBytes": 21691208,
"heapUsedBytes": 103821560,
"domNodes": 13667,
"jsHeapTotalBytes": 11534336,
"scriptDurationMs": 29.47500000000001,
"eventListeners": 2383,
"totalBlockingTimeMs": 127,
"frameDurationMs": 16.670000000000012,
"p95FrameDurationMs": 16.80000000000109
},
{
"name": "viewport-pan-sweep",
"durationMs": 9017.067999999996,
"styleRecalcs": 249,
"styleRecalcDurationMs": 40.758,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 6208.526,
"heapDeltaBytes": -13612764,
"heapUsedBytes": 67569328,
"domNodes": -228,
"jsHeapTotalBytes": 1077248,
"scriptDurationMs": 2871.169,
"eventListeners": -164,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "viewport-pan-sweep",
"durationMs": 8290.081999999984,
"styleRecalcs": 250,
"styleRecalcDurationMs": 36.037,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 5523.471,
"heapDeltaBytes": -323088,
"heapUsedBytes": 80535308,
"domNodes": -228,
"jsHeapTotalBytes": 745472,
"scriptDurationMs": 2649.172,
"eventListeners": -142,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-large-graph-idle",
"durationMs": 17672.627999999917,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 17539.011,
"heapDeltaBytes": -67147800,
"heapUsedBytes": 173684908,
"domNodes": -8049,
"jsHeapTotalBytes": -12783616,
"scriptDurationMs": 280.76,
"eventListeners": -14923,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.80000000000291
},
{
"name": "vue-large-graph-idle",
"durationMs": 17211.609000000066,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 16846.436,
"heapDeltaBytes": -69252988,
"heapUsedBytes": 171585404,
"domNodes": -8049,
"jsHeapTotalBytes": -16191488,
"scriptDurationMs": 265.719,
"eventListeners": -14923,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "vue-large-graph-pan",
"durationMs": 22596.588999999993,
"styleRecalcs": 166,
"styleRecalcDurationMs": 24.895,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 22566.15,
"heapDeltaBytes": -72410660,
"heapUsedBytes": 168531200,
"domNodes": -8049,
"jsHeapTotalBytes": -20910080,
"scriptDurationMs": 936.0870000000001,
"eventListeners": -14923,
"totalBlockingTimeMs": 99,
"frameDurationMs": 17.776666666666763,
"p95FrameDurationMs": 16.799999999995634
},
{
"name": "vue-large-graph-pan",
"durationMs": 21640.35100000001,
"styleRecalcs": 149,
"styleRecalcDurationMs": 19.884000000000015,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 21474.437,
"heapDeltaBytes": -72157880,
"heapUsedBytes": 168843000,
"domNodes": -8049,
"jsHeapTotalBytes": -20910080,
"scriptDurationMs": 935.276,
"eventListeners": -14923,
"totalBlockingTimeMs": 174,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-node-resize-workload",
"durationMs": 6419.901999999979,
"styleRecalcs": 20,
"styleRecalcDurationMs": 72.357,
"layouts": 20,
"layoutDurationMs": 25.661000000000005,
"taskDurationMs": 6403.157999999999,
"heapDeltaBytes": -31209104,
"heapUsedBytes": 196453700,
"domNodes": -8049,
"jsHeapTotalBytes": 12382208,
"scriptDurationMs": 122.359,
"eventListeners": -14923,
"totalBlockingTimeMs": 413,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.80000000000291
},
{
"name": "vue-node-resize-workload",
"durationMs": 6295.424000000025,
"styleRecalcs": 20,
"styleRecalcDurationMs": 71.01899999999998,
"layouts": 20,
"layoutDurationMs": 24.094000000000005,
"taskDurationMs": 6254.170999999999,
"heapDeltaBytes": -47610172,
"heapUsedBytes": 177723580,
"domNodes": -8049,
"jsHeapTotalBytes": 7856128,
"scriptDurationMs": 115.371,
"eventListeners": -14923,
"totalBlockingTimeMs": 455,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "vue-zoom-out-idle",
"durationMs": 8771.008000000053,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 8630.58,
"heapDeltaBytes": -6830420,
"heapUsedBytes": 183904804,
"domNodes": 0,
"jsHeapTotalBytes": -12582912,
"scriptDurationMs": 137.47999999999993,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.780000000000047,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "vue-zoom-out-idle",
"durationMs": 8659.193999999956,
"styleRecalcs": 0,
"styleRecalcDurationMs": 0,
"layouts": 0,
"layoutDurationMs": 0,
"taskDurationMs": 8564.64,
"heapDeltaBytes": 24025688,
"heapUsedBytes": 197432540,
"domNodes": 0,
"jsHeapTotalBytes": -25427968,
"scriptDurationMs": 134.51000000000002,
"eventListeners": 0,
"totalBlockingTimeMs": 0,
"frameDurationMs": 17.776666666666642,
"p95FrameDurationMs": 16.799999999999272
},
{
"name": "workflow-execution",
"durationMs": 106.86900000007427,
"styleRecalcs": 6,
"styleRecalcDurationMs": 13.011999999999999,
"layouts": 3,
"layoutDurationMs": 1.0849999999999997,
"taskDurationMs": 81.419,
"heapDeltaBytes": 3224252,
"heapUsedBytes": 65201576,
"domNodes": 98,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 10.481,
"eventListeners": 33,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
},
{
"name": "workflow-execution",
"durationMs": 474.6099999999842,
"styleRecalcs": 20,
"styleRecalcDurationMs": 19.921999999999997,
"layouts": 4,
"layoutDurationMs": 1.6610000000000003,
"taskDurationMs": 129.00100000000003,
"heapDeltaBytes": 5069144,
"heapUsedBytes": 66959384,
"domNodes": 126,
"jsHeapTotalBytes": 0,
"scriptDurationMs": 11.796999999999999,
"eventListeners": 81,
"totalBlockingTimeMs": 0,
"frameDurationMs": 16.666666666666668,
"p95FrameDurationMs": 16.700000000000728
}
]
}🔬 E2E Coverage
Top 10 files by uncovered lines
🧩 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 |
Matrix rank 79 — "an agent-generated job cannot be reopened as a workflow" (slack-63). Never authored: the stated blocker was that the agent fixture had no execution surface, which this stack supplies. The job under test is a SERVER-submitted run: the agent's `run` tool submits cloud-side, so the browser first learns of the job from the queue, never having POSTed it. - green guard — a completed agent-submitted job whose detail carries the editable graph opens in a new tab holding that graph's nodes. - live-defect pin (test.fail) — when the run was built from an API-format draft, which cloud can produce and cannot convert, `openJobWorkflow` hits `if (!data) return` and the click produces nothing at all: no tab, no toast, no dialog. The pin is on the user being told, not on the tab opening. Adds `AgentExecutionHelper.mockJobDetail`, a pass-through to the existing `AssetsHelper.mockJobDetail`, so a spec about what the action does with the job detail supplies one rather than reading a failed fetch as a product verdict. No production code changes.
d6e4495 to
f2000ea
Compare
Playwright videos for new testsDownload the recorded videos and HTML report Recorded projects: cloud Recorded test titles:
|
Codecov Report✅ All modified and coverable lines are covered by tests. @@ Coverage Diff @@
## christian-byrne/qspec5-execution-story-specs #18936 +/- ##
================================================================================
- Coverage 84.89% 84.88% -0.01%
================================================================================
Files 2303 2303
Lines 132215 132214 -1
Branches 33778 35684 +1906
================================================================================
- Hits 112241 112232 -9
- Misses 19676 19684 +8
Partials 298 298
Flags with carried forward coverage won't be shown. Click here to find out more. 🚀 New features to boost your workflow:
|
54d7c4e to
a7796b7
Compare
35c265c to
58fc8a1
Compare
da62d72 to
b731bf3
Compare
…#18936) Consolidates the one spec #18936 contributed that is not already on `main` or carried here. Everything else on that branch is a duplicate: its harness commit merged as #18790 (`d90da45a1b`), its `agentExecutionLifecycle.spec.ts` is on `main` with the same three case titles, and its `agentGoldenPathRunIt` and `agentStopKeepsServerRunAlive` are earlier, weaker drafts of this branch's. #18936's base is this branch, 1054 commits behind `main`. Two adaptations rather than a verbatim copy: - `enqueueServerRun` is called with the signature `main` actually has. #18936 added an optional `{ workflowId }` to the shared helper to stamp `workflow_id` on the list row; the queue's `open-workflow` menu entry is unconditional for a completed job (`useJobMenu.ts`, `jobMenuEntries`), so the field bought nothing and widening a shared harness for it is not warranted. - The expected-failure window is narrowed from the whole test to the single assertion it is about. #18936 wrapped the case in `test.fail(title, fn)`, which absorbs *any* throw as the expected PM defect — a job that never reaches the queue, a "More" menu that never opens, a missing menu entry, or a tab that unexpectedly opens would all have counted as a pass. `test.fail` is now declared immediately before the toast assertion, and the "no tab opened" check sits outside that window, so the pin fails loudly if anything other than the named defect is what happens. The defect is live on `main` at `12254f251c`: `useJobMenu.openJobWorkflow` reads `getJobWorkflow(target.id)` and `if (!data) return`, so a job stored with no `extra_pnginfo.workflow` — which is what an agent-built API-format draft produces — closes the menu and says nothing at all. The pin asserts a visible refusal and deliberately does not require the tab to open. `mockJobDetail` is a passthrough to the existing `AssetsHelper.mockJobDetail`; without a `GET /api/jobs/{id}` route the fetch 404s and the spec would read a failed request as a product verdict.
|
Closing — folded into #18803, which is the surviving qspec5 carrier. The one spec here that is not already on Why the rest is not folded, verified file-by-file against
This branch bases on #18803s branch at a point 1054 commits behind Two deliberate changes were made to the spec rather than copying it verbatim, both recorded in the fold commit message:
The defect the pin names is still live on |
…re/stall surfaced, finished-generation preview (Comfy-Org#18803) ## Summary Adds black-box user-story specs on the agent harness execution surface (Comfy-Org#18790), stacked on the rank-26 golden path (Comfy-Org#18802), and extends the golden path with the finished run's history-row thumbnail. ## Changes - **What**: - `agentStopKeepsServerRunAlive.spec.ts` (rank 41): Stop cancels the agent's turn but sends no job interrupt or cancel request, and the submitted job keeps its executing outline. The story fixture records `/api/interrupt`, `/api/jobs/cancel` and `/api/jobs/{id}/cancel` as `jobCancelRequests()`, and accepts the turn-cancel request. - `agentFailedRunIsSurfaced.spec.ts` (rank 99, failure half): a server-submitted run that fails after the turn ends shows its error message in the queue row's details popover, and the agent panel stays idle. - `agentGoldenPathRunIt.spec.ts` (rank 98, fixture-supported half): the finished run's history row renders a thumbnail served from that run's output. - `agentJobReopenedAsWorkflow.spec.ts` (rank 79, folds Comfy-Org#18936): "open as workflow" rebuilds the graph when the job carries one, and — as a declared expected failure — says nothing at all when it does not. - All four live beside their fixture in `browser_tests/tests/agent/`. ## Review Focus - **Why the fixture moved.** The fixture and its execution helper additions are used only by the four specs in this PR, so they now live next to them in `browser_tests/tests/agent/` instead of in `browser_tests/fixtures/` — the same shape as `agentPanelMocks.ts` and `agentCreditsExhaustionLifecycleMocks.ts`, which are already `agentTest.extend` fixtures under `tests/agent/`. CODEOWNERS already scoped both files to `@Comfy-Org/agent_devs` through `/browser_tests/fixtures/**/*gent*`, so this does not change who owns them; it only makes the paths agree with that ownership. The branch ruleset's narrower `browser_tests/fixtures/data/agent/**` glob did not cover them, which is why the earlier layout pulled an Agent-test-only PR into the shared-path reviewer entry. - The two story-only additions to the execution surface — the job interrupt/cancel recorder and a per-job `GET /api/jobs/{id}` detail route — are now a subclass in the story fixture rather than changes to the shared `AgentExecutionHelper`, which `agentExecutionFixture.ts` and `agentExecutionLifecycle.spec.ts` also consume. This PR no longer modifies that shared helper. - The stalled-run half of rank 99 and rank 42 (output on the producing node) are not separate specs. `agentExecutionLifecycle.spec.ts` and the stop spec already cover them with the same steps. - The rank-98 agent-panel half (the reply presenting a result link) is deferred until a recorded conversation or a product contract for result links exists. - Product gap, not asserted: for a server-submitted run, `execution_error` does not open the error dialog, because `executionStore.runErrorKeyForJob` cannot resolve a workflow path for the job and buffers the error as unattributable. The queue row details are the surface that does show it. - Verified locally against a cloud dev server at the current head: all four specs pass, with the rank-79 second case failing exactly where it declares it will. The stop spec's empty-cancel-list assertion is proven non-vacuous two ways — mutating Stop to also call `api.interrupt` fails it, and a positive control that issues `/api/interrupt` from the page records it. --------- Co-authored-by: Alexander Brown <drjkl@comfy.org>
Summary
Black-box coverage for "an agent-generated job cannot be reopened as a workflow" (user-story matrix rank 79, slack-63). Tests only; no production code changes.
The job under test is a server-submitted run: the agent's
runtool submits cloud-side, so the browser first learns of the job from the queue, never having POSTed it (submittedPrompts()is asserted empty). That is what makes this a story about agent jobs rather than about the queue in general.Changes
browser_tests/tests/agent/agentJobReopenedAsWorkflow.spec.ts(1 green guard, 1 live-defect pin), plusAgentExecutionHelper.mockJobDetail— a pass-through to the existingAssetsHelper.mockJobDetail.test.fail, red onmain) — when the detail carries a runnable prompt but no editable graph, the click produces nothing at all.Review Focus
Why the second case is realistic, and not a mock I invented to fail.
openJobWorkflow(useJobMenu.ts) readsGET /api/jobs/{id}and pulls the graph out ofworkflow.extra_data.extra_pnginfo.workflow. Cloud stores thatworkflowas the verbatim/api/promptbody (services/ingest/server/implementation/prompt.go,SetWorkflowJSON), so what the action can do is decided entirely by what the submitter sent — and an agent that built the graph itself can send an API-format draft. Cloud says so in its own words (services/agent/internal/loop/tools.go): such a draft "runs but isn't yet editable on the canvas", there is no API→frontend converter in comfy-cli, anddraftNeedsFrontendConversionexists precisely so the condition is observable rather than silent.What the pin actually claims. Not that the tab must open — rebuilding a canvas from an API prompt is real work and may well belong server-side. Only that the user must be told. Today:
…so the click does nothing: no tab, no toast, no dialog. The message the original report quotes, "no workflow data is available", is
mediaAsset.noWorkflowDataFound— the asset route's copy (useMediaAssetActions.ts). The job-menu route says nothing at all. The pin asserts a toast, because that is how the asset route refuses the same action; it is the mechanism this route is missing, not an invented one.A second symptom I found and deliberately did not assert: the context menu does not close either, so the click leaves the UI byte-for-byte as it was. Keeping the pin to one claim keeps it honest about which fix satisfies it. Worth a look while fixing.
Why the job detail is mocked. Without a
/api/jobs/{id}route that fetch simply 404s, and a spec about what the action does with the detail would be reading a failed fetch as a product verdict. Both cases supply a detail; only theworkflowfield differs between them, which is what makes them each other's mutation check.Surface note. The queue overlay is used rather than the Job History sidebar tab (the harness pins
Comfy.Queue.QPOV2: false). Both render the sameuseJobMenuentry and call the sameopenJobWorkflow, so the coverage is not surface-specific.Verification
Local cloud rig (own ComfyUI e2e container, dev server on the 6200 band),
--project=cloud --retries=0.--repeat-each=3, zero flake.test.failand reading the failure. An earlier, looser refusal locator matched the job menu's own popover (role="dialog"); that false positive is why the assertion is now narrowed to a toast.agentExecutionLifecycle,agentRunOutputReachesSaveNode,agentStopKeepsServerRunAlive(5/5).typecheck:browser,oxfmt --check,oxlint --type-aware,eslintclean; pre-commit and pre-push (knip, fallow) pass,complexity_introduced: 0.