[workflow] Load per-source replay bundles

[NathanColosimo]

High Level

Change build output to be separate workflow bundles and load only the needed code in the runtime.
There's very few changes to the runtime code, most of the changes are in the builder code.
Almost all of the changes are tests

Summary

• emit one VM bundle per workflow source; workflows in the same source share one bundle
• keep one flow route with a static workflow-ID → dynamic-import loader map
• overlap the selected bundle load with world initialization and event setup
• skip bundle loading for queue-only step deliveries
• retain compiled scripts for every immutable production source bundle
• keep watch loaders cache-safe and derive loader IDs from the same transform that emitted each bundle
• encode generated VM modules as opaque artifacts so framework plugins cannot rewrite inert source

Result

For the Next/Turbopack benchmark app's 97_bench source, measured against the monolithic bundle at the base of this PR:

• generated flow route: 1,327,514 B → 18,264 B (-98.6%)
• selected benchmark VM code: 1,307,595 B → 133,859 B (-89.8%)
• cold compile p50: 10.405 ms → 1.089 ms (-89.5%)
• fresh-context evaluation p50: 7.484 ms → 0.356 ms (-95.2%)
• one-time base64 artifact decode p50/p90: 0.037/0.044 ms

The compile/evaluation microbenchmarks use 60 uncached vm.Script compilations and 80 evaluations in fresh VM contexts on Node. The deployment benchmark and APM traces remain the end-to-end signal.

The correctness-preserving split deliberately includes any source that defines both workflows and custom serializers in every other source bundle. The workbench's unusually large 99_e2e.ts is such a hybrid, so the selected benchmark bundle is ~134 KB rather than the ~28 KB it would be without that shared registration. The next structural improvement is to extract serializer registration into a dedicated small VM prelude; omitting it would make replay deserialization incorrect.

The 19 split bundles total 5.28 MB decoded versus the old 1.31 MB monolith because shared sandbox dependencies and the hybrid serializer source repeat. Opaque base64 transport is 7.04 MB on disk before compression (1.33× decoded; gzip is close to the original text). This is a build/storage tradeoff: a replay imports, decodes, compiles, and evaluates only its selected source bundle, and the route caches that decoded promise.

Deployment benchmark

The final benchmark run completed successfully on deployment dpl_CHJqdRSi52QvbKm1KzXhEs4Aa1B2; the sticky comparison reports:

• inline STSO p75 191 → 169 ms (-12%), p90 229 → 191 ms (-17%), p99 580 → 268 ms (-54%)
• whole-run overhead for 1,020 steps 195,405 → 167,186 ms (-14%)
• an earlier run of the same runtime code measured STSO p75/p90/p99 at 155/172/264 ms and whole-run overhead at 153,275 ms, so the improvement survives expected cold-start and infrastructure variance
• TTFS p75 remains 1.12–1.50 s across scenarios, showing that dispatch, route cold start, and workflow-server work still dominate end-to-end startup

A final cold compile-miss trace has a 22.52 ms workflow.run; the surrounding fresh-replay path shows 6.88 ms bundle load (overlapped with 2.37 ms world init), followed by 3.00 ms context creation, 4.00 ms compile, 2.63 ms evaluate, 1.53 ms input hydrate, and 8.82 ms replay execute. Its 146 ms payload-preparation wall span begins before the replay and overlaps event loading and execution rather than representing 146 ms of blocking CPU work.

After the bundle and script caches are warm, the final sequential benchmark's first fresh VM replay trace is 6.88 ms: context creation 2.00 ms, compile hit 0.14 ms, evaluation 1.94 ms, input hydration 0.12 ms, and replay execution 2.41 ms. The comparable main trace is 186.10 ms (-96.3% for this representative warm-cache replay). Subsequent retained invocations in the final trace are generally 1–2.4 ms.

Validation

• 295 focused tests across builders, runtime, tracing, payload cache, and stream recovery
• complete builders suite: 226/226
• affected package typechecks and builds
• Next/Turbopack, Vite, and Nuxt production builds
• exact Next HMR regression plus same-size/same-mtime workflow-ID rename regression
• all eight stable/canary × Webpack/Turbopack × Node/QuickJS Next dev E2E lanes green
• Linux and Windows unit lanes plus Node/QuickJS Windows Next E2E lanes green
• standalone/BOPA import smoke test
• final benchmark workflow green
• final Codex and Claude autoreviews clean

[changeset-bot]

🦋 Changeset detected

Latest commit: 770e721

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 16 packages

Name	Type
@workflow/builders	Patch
@workflow/core	Patch
@workflow/next	Patch
@workflow/world-testing	Patch
workflow	Patch
@workflow/astro	Patch
@workflow/cli	Patch
@workflow/nest	Patch
@workflow/nitro	Patch
@workflow/rollup	Patch
@workflow/sveltekit	Patch
@workflow/vite	Patch
@workflow/vitest	Patch
@workflow/web-shared	Patch
@workflow/web	Patch
@workflow/nuxt	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-express-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nestjs-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-python-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-docs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-swc-playground	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-tarballs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-web	Ready	Preview, v0	Sep 1, 2026 3:32am UTC

[github-actions]

🧪 E2E Test Results

❌ Some tests failed

❌ Failed E2E Tests

▲ Vercel Production (8 failed)

python-node (8 failed):

• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3 | 🔍 observability
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M | 🔍 observability
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23 | 🔍 observability
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG | 🔍 observability
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115 | 🔍 observability
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP | 🔍 observability
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY | 🔍 observability
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF | 🔍 observability

🌐 Cross-language Conformance (9 failed)

python (9 failed):

• deploymentId: 'latest' is a no-op in non-Vercel worlds | wrun_01M1DGN5VWSB6FR072SCMVKD93
• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

• cancelRun via CLI - cancelling a running workflow (nextjs-turbopack)
• cancelRun via CLI - cancelling a running workflow (nuxt)
• promiseAllWorkflow (nest)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

31 infra events

• cold-start-warmup · suite warmup (python) · at 03:32:38Z · abandoned wrun_41M1DG9NKP0GT49DFD2FT24FQP · (+7 more)
• run-pickup-stall · nullByteWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA4A0GWNBVG1J3PNW2BM
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F7
• run-pickup-stall · sleepingWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F6
• run-pickup-stall · promiseAllWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA410GG3J9DQS5J8V5RX
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDACG0GYN2PV31FE1722G
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:33:26Z · abandoned wrun_41M1DGE9JE0GYK5Q2XK9JWVR87
• run-pickup-stall · sleepingWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5DP0GZT0NGX0185G3X8
• run-pickup-stall · promiseAllWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5BE0GMZ14CQP9SC1Y7K
• run-pickup-stall · nullByteWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5AD0GMMXJ2AJ6AZN3XB
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:33:55Z · abandoned wrun_41M1DGF5BM0GJKW6R6A11PP5G5
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:34:45Z · abandoned wrun_41M1DGF8SE0GTQDHMN40HTGC3X
• cold-start-warmup · suite warmup (python) · at 03:35:40Z · abandoned wrun_01M1DGF7EGPY04C8PJ02V841CN · (+7 more)
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJV4MD7KSP260W7ZSBC
• run-pickup-stall · sleepingWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK0G2QC0ZT6CH9PZYC0
• run-pickup-stall · promiseAllWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJW8G1Y4FFC31VK77KY
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK38F43AA3AECZXYP9H
• run-pickup-stall · nullByteWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK5160SZJ5HM2H6FMB9
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:36:26Z · abandoned wrun_41M1DGH6TB0GZCNX8E7RCMJJ6N
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:36:26Z · abandoned wrun_41M1DGJ0P00GPSJ9R5XE1S1WHX
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6HAQR493S719104TSK
• run-pickup-stall · sleepingWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6NBAJ7B203HQDK5N2N
• run-pickup-stall · promiseAllWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6KR8J5R749MV6Q056B
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6SZKHYHZ67N6HHVCTV
• run-pickup-stall · nullByteWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ70B63HA9BG628J3E8C
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHT45WP6FSWB7FGZRVDY
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTEREXVJRZ204H80RN5
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTKKDTFSEZV6GQKNE3R
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF60XCRG4QP1WMRPCM2K
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF5YJJWK6M7RQPW03HKJ
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:38:55Z · abandoned wrun_01M1DGRCDP4DNKXYYTR30VTRPT

E2E Test Summary

Summary

	Passed	Failed	Skipped	Total
❌ ▲ Vercel Production	3570	8	742	4320
✅ 💻 Local Development	3922	0	558	4480
✅ 📦 Local Production	3922	0	558	4480
✅ 🐘 Local Postgres	3922	0	558	4480
✅ 🪟 Windows	320	0	0	320
❌ 🌐 Cross-language Conformance	0	9	132	141
✅ vercel-http-transport	817	0	143	960
✅ vercel-multi-region	27	0	0	27
✅ vercel-ws-transport	553	0	87	640
Total	17053	17	2778	19848

Details by Category

❌ ▲ Vercel Production

App	Passed	Failed	Skipped
✅ astro-node	132	0	28
✅ astro-quickjs	132	0	28
✅ example-node	132	0	28
✅ example-quickjs	132	0	28
✅ express-node	132	0	28
✅ express-quickjs	132	0	28
✅ fastify-node	132	0	28
✅ fastify-quickjs	132	0	28
✅ hono-node	132	0	28
✅ hono-quickjs	132	0	28
✅ nest-node	132	0	28
✅ nest-quickjs	132	0	28
✅ nextjs-turbopack-node	157	0	3
✅ nextjs-turbopack-quickjs	157	0	3
✅ nextjs-webpack-node	157	0	3
✅ nextjs-webpack-quickjs	157	0	3
✅ nitro-node	132	0	28
✅ nitro-quickjs	132	0	28
✅ nuxt-node	132	0	28
✅ nuxt-quickjs	132	0	28
❌ python-node	0	8	152
✅ sveltekit-node	151	0	9
✅ sveltekit-quickjs	151	0	9
✅ tanstack-start-node	132	0	28
✅ tanstack-start-quickjs	132	0	28
✅ vite-node	132	0	28
✅ vite-quickjs	132	0	28

✅ 💻 Local Development

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 📦 Local Production

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🐘 Local Postgres

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	160	0	0
✅ nextjs-turbopack-quickjs	160	0	0

❌ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
❌ python	0	9	132

✅ vercel-http-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ hono	132	0	28
✅ nextjs-turbopack	157	0	3
✅ nitro	132	0	28
✅ vite	132	0	28

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ nextjs-turbopack	157	0	3
✅ vite	132	0	28

---

📋 View full workflow run

[github-actions]

📊 Workflow Benchmarks

commit 770e721 · Tue, 01 Sep 2026 03:46:20 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	1102 (+502%) 🔻	1224 🔴 (+15%) 🔻	1258 🔴 (+13%)	1306 🔴 (-8.5%)	30
TTFS	stream	112 (-48%) 💚	1187 🔴 (+9.2%)	1225 🔴 (+12%)	1243 🔴 (+6.9%)	30
TTFS	hook + stream	377 (-69%) 💚	1463 🔴 (+9.8%)	1836 🔴 (+32%) 🔻	2132 🔴 (+36%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	508 (-15%) 💚	1679 (+5.7%)	1705 (-3.1%)	1752 (-6.7%)	10
Fan-out TTLS	Promise.all(100 steps)	1590 (-4.6%)	3569 (+6.8%)	3692 (+4.0%)	4409 (-34%) 💚	10
STSO	1020 steps (inline)	67 (-40%) 💚	133 (-10%)	155 (-11%)	230 (-35%) 💚	1019
WO	1020 steps	131497 (-12%)	131497 (-12%)	131497 (-12%)	131497 (-12%)	1
CRTT	first chunk (pooled)	88 (-9.3%)	120 (-20%) 💚	178 (+3.5%)	355 (+39%) 🔻	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	110 (-25%)	120 (-33%)	185 (-52%)	315 (-58%)	116 (-40%)	10
size sweep (100/s, 160B-12KB)	106 (-17%)	132 (-29%)	268 (-21%)	767 (+62%)	133 (-13%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	117 (-11%)	110 (-31%)	130 (-48%)	224 (-63%)	251 (-24%)	3
replay eve-gpt-5.6-sol-2000t (1x)	115 (-32%)	121 (-29%)	150 (-37%)	335 (-61%)	392 (-36%)	2
replay eve-gpt-5.6-sol-2000t (2x)	120 (-6%)	149 (-38%)	225 (-36%)	457 (-28%)	254 (-11%)	3

📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 148790ms → this run 131259ms (Δ -17531ms, -12%)

 50-100 ms ┃ main 0 this 3 +3
100-150 ms █████████████████████░░┃ main 779 this 872 +93
150-200 ms ███┃█ main 181 this 128 -53
200-250 ms ┃ main 31 this 9 -22
250-300 ms ┃ main 11 this 6 -5
300-350 ms ┃ main 6 this 0 -6
350-400 ms ┃ main 6 this 0 -6
400-450 ms ┃ main 1 this 1 +0
450-500 ms ┃ main 1 this 0 -1
500-550 ms ┃ main 1 this 0 -1
600-650 ms ┃ main 1 this 0 -1
650-700 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)

variant RTT 1ms→5s+ avg p50 p90 p99 n
control ·····▁█▇▁···· 103.2 (-32%) 96 (-24%) 185 (-52%) 315 (-58%) 3000
sweep ······▆█▁▁··· 126.4 (-17%) 107 (-15%) 268 (-21%) 767 (+62%) 3000
gw 1x ·····▁█▅▁···· 95.8 (-30%) 92 (-23%) 130 (-48%) 224 (-63%) 5295
eve 1x ·····▁█▇▁▁··· 103.6 (-35%) 94 (-27%) 150 (-37%) 335 (-61%) 5186
eve 2x ·····▁▅█▂▁··· 128.1 (-32%) 113 (-31%) 225 (-36%) 457 (-28%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▂▃▂▁▅▂▁▃▃ 95–124ms
sweep █▅▃▂▂▄▂▄▂▁ 105–169ms
gw 1x ▃█▃▄▂▂▂▁▂▄ 91–107ms
eve 1x ▃▂▃▃▁█▃▄▂▁ 86–146ms
eve 2x ▂▃▃▁▁▃▆█▅▂ 103–175ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep ▅█▆▃▂▁▂ 123–131ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▅▃▄█▃▅▅▁▃▆ 26–35ms
sweep ▅▆▇▄▇▄▆█▄▁ 41–56ms
gw 1x ▁█▃▃▃▄▆▅▄▅ 23–32ms
eve 1x ▂▃▁▁▂█▅▅▃▃ 19–29ms
eve 2x ▂▄▁▁▁▂█▄▂▆ 19–26ms

ℹ️ Metric definitions & methodology

_{Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.}

_{The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.}

_{The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.}

_{Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.}

_{Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)}

_{Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles}

_{Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600}

_{All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).}

_{Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.}

[github-actions]

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenario	outcome	events	virt	replay	violations
✅ smoke-no-steps	completed	3	0ms	ok	0
✅ smoke-one-step	completed	6	0ms	ok	0
✅ hook-at-step-started	completed	12	0ms	ok	0
✅ hook-at-step-completed	completed	12	0ms	ok	0
✅ hook-at-hook-created	completed	12	0ms	ok	0
✅ deadline-hook-wins	completed	7	1.0h	ok	0
✅ deadline-expires	completed	7	1.0h	ok	0
✅ long-sleep	completed	11	30.0d	ok	0
✅ hook-never-arrives	stalled	3	0ms	skipped	0
✅ step-retries-twice	completed	10	2.0s	ok	0
✅ parallel-steps	completed	9	0ms	ok	0
✅ hook-on-execution-state	completed	12	0ms	ok	0
✅ peek-hook-before-branch	completed	12	0ms	ok	0
✅ peek-hook-after-branch	completed	12	0ms	ok	0
✅ peek-hook-at-registration	completed	12	0ms	ok	0
✅ race-hook-before-probe	completed	12	0ms	ok	0
✅ race-hook-after-probe	completed	12	0ms	ok	0
✅ race-duplicate-delivery	completed	13	0ms	ok	0
✅ attr-hook-before-step	completed	11	0ms	ok	0
✅ attr-hook-after-step	completed	11	0ms	ok	0
✅ attr-from-step-body	completed	13	0ms	ok	0
✅ fork-hook-after-timeout	completed	14	1.0m	ok	0
✅ fork-hook-before-timeout	completed	14	1.0m	ok	0
✅ count-hook-after-timeout	completed	17	1.0m	ok	0
✅ count-hook-before-timeout	completed	20	1.0m	ok	0
✅ stale-read-step-count-fork	completed	20	1.0m	ok	0
✅ stale-read-equal-step-counts	completed	14	1.0m	ok	0
✅ step-vs-step-fork	completed	12	0ms	ok	0
✅ step-vs-step-fork-fenced	completed	12	0ms	ok	0
✅ fence-catches-benign-direction	completed	12	5ms	ok	0
✅ in-flight-before-decision	completed	17	1.0m	ok	0
❌ in-flight-before-decision-counted	completed	17	1.0m	ok	0
✅ in-flight-after-decision	completed	19	2.0m	ok	0
✅ stale-read-step-count-fork-fenced	completed	20	1.0m	ok	0
✅ fork-hook-wins	completed	13	1.0m	ok	0
✅ fork-timeout-wins	completed	13	1.0m	ok	0
✅ unclaimed-payload-under-fork	completed	17	1.0m	ok	0
✅ claimed-payload-under-fork	completed	17	1.0m	ok	0
✅ writers-independent-step-bodies	completed	12	0ms	ok	0
✅ writers-scripted-tempo	completed	12	0ms	ok	0
✅ cancel-mid-step	cancelled	7	0ms	skipped	0

Full trace: world-sim.txt

[vercel]

Additional Suggestion:

Test asserts SPEC_VERSION_CURRENT === SPEC_VERSION_SUPPORTS_COMPRESSION (5) but the PR bumped SPEC_VERSION_CURRENT to SPEC_VERSION_SUPPORTS_SLOT_IDENTITY (6), so the assertion evaluates 6 === 5 and fails.

[github-actions]

Framework	Cold replay	Step reg.	Framework output
hono	304.7 KiB (+103.8 KiB) ⚠️	40.6 KiB (±0)	2.58 MiB (+830.6 KiB)
nextjs-turbopack	305.0 KiB (+98.7 KiB) ⚠️	439 B (±0)	562.1 KiB (-203.4 KiB)

About these numbers

Sizes are gzip; parentheses show the change against main.
Cold replay and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.
⚠️ marks growth past the threshold. Add the allow-bundle-size-growth label to accept it.

770e721 · run

[NathanColosimo]

High Level

Change build output to be separate workflow bundles and load only the needed code in the runtime.
There's very few changes to the runtime code, most of the changes are in the builder code.
Almost all of the changes are tests

Summary

• emit one VM bundle per workflow source; workflows in the same source share one bundle
• keep one flow route with a static workflow-ID → dynamic-import loader map
• overlap the selected bundle load with world initialization and event setup
• skip bundle loading for queue-only step deliveries
• retain compiled scripts for every immutable production source bundle
• keep watch loaders cache-safe and derive loader IDs from the same transform that emitted each bundle
• encode generated VM modules as opaque artifacts so framework plugins cannot rewrite inert source

Result

For the Next/Turbopack benchmark app's 97_bench source, measured against the monolithic bundle at the base of this PR:

• generated flow route: 1,327,514 B → 18,264 B (-98.6%)
• selected benchmark VM code: 1,307,595 B → 133,859 B (-89.8%)
• cold compile p50: 10.405 ms → 1.089 ms (-89.5%)
• fresh-context evaluation p50: 7.484 ms → 0.356 ms (-95.2%)
• one-time base64 artifact decode p50/p90: 0.037/0.044 ms

The compile/evaluation microbenchmarks use 60 uncached vm.Script compilations and 80 evaluations in fresh VM contexts on Node. The deployment benchmark and APM traces remain the end-to-end signal.

The correctness-preserving split deliberately includes any source that defines both workflows and custom serializers in every other source bundle. The workbench's unusually large 99_e2e.ts is such a hybrid, so the selected benchmark bundle is ~134 KB rather than the ~28 KB it would be without that shared registration. The next structural improvement is to extract serializer registration into a dedicated small VM prelude; omitting it would make replay deserialization incorrect.

The 19 split bundles total 5.28 MB decoded versus the old 1.31 MB monolith because shared sandbox dependencies and the hybrid serializer source repeat. Opaque base64 transport is 7.04 MB on disk before compression (1.33× decoded; gzip is close to the original text). This is a build/storage tradeoff: a replay imports, decodes, compiles, and evaluates only its selected source bundle, and the route caches that decoded promise.

Deployment benchmark

The final benchmark run completed successfully on deployment dpl_CHJqdRSi52QvbKm1KzXhEs4Aa1B2; the sticky comparison reports:

• inline STSO p75 191 → 169 ms (-12%), p90 229 → 191 ms (-17%), p99 580 → 268 ms (-54%)
• whole-run overhead for 1,020 steps 195,405 → 167,186 ms (-14%)
• an earlier run of the same runtime code measured STSO p75/p90/p99 at 155/172/264 ms and whole-run overhead at 153,275 ms, so the improvement survives expected cold-start and infrastructure variance
• TTFS p75 remains 1.12–1.50 s across scenarios, showing that dispatch, route cold start, and workflow-server work still dominate end-to-end startup

A final cold compile-miss trace has a 22.52 ms workflow.run; the surrounding fresh-replay path shows 6.88 ms bundle load (overlapped with 2.37 ms world init), followed by 3.00 ms context creation, 4.00 ms compile, 2.63 ms evaluate, 1.53 ms input hydrate, and 8.82 ms replay execute. Its 146 ms payload-preparation wall span begins before the replay and overlaps event loading and execution rather than representing 146 ms of blocking CPU work.

After the bundle and script caches are warm, the final sequential benchmark's first fresh VM replay trace is 6.88 ms: context creation 2.00 ms, compile hit 0.14 ms, evaluation 1.94 ms, input hydration 0.12 ms, and replay execution 2.41 ms. The comparable main trace is 186.10 ms (-96.3% for this representative warm-cache replay). Subsequent retained invocations in the final trace are generally 1–2.4 ms.

Validation

• 295 focused tests across builders, runtime, tracing, payload cache, and stream recovery
• complete builders suite: 226/226
• affected package typechecks and builds
• Next/Turbopack, Vite, and Nuxt production builds
• exact Next HMR regression plus same-size/same-mtime workflow-ID rename regression
• all eight stable/canary × Webpack/Turbopack × Node/QuickJS Next dev E2E lanes green
• Linux and Windows unit lanes plus Node/QuickJS Windows Next E2E lanes green
• standalone/BOPA import smoke test
• final benchmark workflow green
• final Codex and Claude autoreviews clean

[changeset-bot]

🦋 Changeset detected

Latest commit: 770e721

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 16 packages

Name	Type
@workflow/builders	Patch
@workflow/core	Patch
@workflow/next	Patch
@workflow/world-testing	Patch
workflow	Patch
@workflow/astro	Patch
@workflow/cli	Patch
@workflow/nest	Patch
@workflow/nitro	Patch
@workflow/rollup	Patch
@workflow/sveltekit	Patch
@workflow/vite	Patch
@workflow/vitest	Patch
@workflow/web-shared	Patch
@workflow/web	Patch
@workflow/nuxt	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-express-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nestjs-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-python-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-docs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-swc-playground	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-tarballs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-web	Ready	Preview, v0	Sep 1, 2026 3:32am UTC

[github-actions]

🧪 E2E Test Results

❌ Some tests failed

❌ Failed E2E Tests

▲ Vercel Production (8 failed)

python-node (8 failed):

• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3 | 🔍 observability
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M | 🔍 observability
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23 | 🔍 observability
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG | 🔍 observability
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115 | 🔍 observability
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP | 🔍 observability
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY | 🔍 observability
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF | 🔍 observability

🌐 Cross-language Conformance (9 failed)

python (9 failed):

• deploymentId: 'latest' is a no-op in non-Vercel worlds | wrun_01M1DGN5VWSB6FR072SCMVKD93
• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

• cancelRun via CLI - cancelling a running workflow (nextjs-turbopack)
• cancelRun via CLI - cancelling a running workflow (nuxt)
• promiseAllWorkflow (nest)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

31 infra events

• cold-start-warmup · suite warmup (python) · at 03:32:38Z · abandoned wrun_41M1DG9NKP0GT49DFD2FT24FQP · (+7 more)
• run-pickup-stall · nullByteWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA4A0GWNBVG1J3PNW2BM
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F7
• run-pickup-stall · sleepingWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F6
• run-pickup-stall · promiseAllWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA410GG3J9DQS5J8V5RX
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDACG0GYN2PV31FE1722G
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:33:26Z · abandoned wrun_41M1DGE9JE0GYK5Q2XK9JWVR87
• run-pickup-stall · sleepingWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5DP0GZT0NGX0185G3X8
• run-pickup-stall · promiseAllWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5BE0GMZ14CQP9SC1Y7K
• run-pickup-stall · nullByteWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5AD0GMMXJ2AJ6AZN3XB
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:33:55Z · abandoned wrun_41M1DGF5BM0GJKW6R6A11PP5G5
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:34:45Z · abandoned wrun_41M1DGF8SE0GTQDHMN40HTGC3X
• cold-start-warmup · suite warmup (python) · at 03:35:40Z · abandoned wrun_01M1DGF7EGPY04C8PJ02V841CN · (+7 more)
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJV4MD7KSP260W7ZSBC
• run-pickup-stall · sleepingWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK0G2QC0ZT6CH9PZYC0
• run-pickup-stall · promiseAllWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJW8G1Y4FFC31VK77KY
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK38F43AA3AECZXYP9H
• run-pickup-stall · nullByteWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK5160SZJ5HM2H6FMB9
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:36:26Z · abandoned wrun_41M1DGH6TB0GZCNX8E7RCMJJ6N
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:36:26Z · abandoned wrun_41M1DGJ0P00GPSJ9R5XE1S1WHX
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6HAQR493S719104TSK
• run-pickup-stall · sleepingWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6NBAJ7B203HQDK5N2N
• run-pickup-stall · promiseAllWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6KR8J5R749MV6Q056B
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6SZKHYHZ67N6HHVCTV
• run-pickup-stall · nullByteWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ70B63HA9BG628J3E8C
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHT45WP6FSWB7FGZRVDY
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTEREXVJRZ204H80RN5
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTKKDTFSEZV6GQKNE3R
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF60XCRG4QP1WMRPCM2K
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF5YJJWK6M7RQPW03HKJ
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:38:55Z · abandoned wrun_01M1DGRCDP4DNKXYYTR30VTRPT

E2E Test Summary

Summary

	Passed	Failed	Skipped	Total
❌ ▲ Vercel Production	3570	8	742	4320
✅ 💻 Local Development	3922	0	558	4480
✅ 📦 Local Production	3922	0	558	4480
✅ 🐘 Local Postgres	3922	0	558	4480
✅ 🪟 Windows	320	0	0	320
❌ 🌐 Cross-language Conformance	0	9	132	141
✅ vercel-http-transport	817	0	143	960
✅ vercel-multi-region	27	0	0	27
✅ vercel-ws-transport	553	0	87	640
Total	17053	17	2778	19848

Details by Category

❌ ▲ Vercel Production

App	Passed	Failed	Skipped
✅ astro-node	132	0	28
✅ astro-quickjs	132	0	28
✅ example-node	132	0	28
✅ example-quickjs	132	0	28
✅ express-node	132	0	28
✅ express-quickjs	132	0	28
✅ fastify-node	132	0	28
✅ fastify-quickjs	132	0	28
✅ hono-node	132	0	28
✅ hono-quickjs	132	0	28
✅ nest-node	132	0	28
✅ nest-quickjs	132	0	28
✅ nextjs-turbopack-node	157	0	3
✅ nextjs-turbopack-quickjs	157	0	3
✅ nextjs-webpack-node	157	0	3
✅ nextjs-webpack-quickjs	157	0	3
✅ nitro-node	132	0	28
✅ nitro-quickjs	132	0	28
✅ nuxt-node	132	0	28
✅ nuxt-quickjs	132	0	28
❌ python-node	0	8	152
✅ sveltekit-node	151	0	9
✅ sveltekit-quickjs	151	0	9
✅ tanstack-start-node	132	0	28
✅ tanstack-start-quickjs	132	0	28
✅ vite-node	132	0	28
✅ vite-quickjs	132	0	28

✅ 💻 Local Development

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 📦 Local Production

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🐘 Local Postgres

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	160	0	0
✅ nextjs-turbopack-quickjs	160	0	0

❌ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
❌ python	0	9	132

✅ vercel-http-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ hono	132	0	28
✅ nextjs-turbopack	157	0	3
✅ nitro	132	0	28
✅ vite	132	0	28

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ nextjs-turbopack	157	0	3
✅ vite	132	0	28

---

📋 View full workflow run

[github-actions]

📊 Workflow Benchmarks

commit 770e721 · Tue, 01 Sep 2026 03:46:20 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	1102 (+502%) 🔻	1224 🔴 (+15%) 🔻	1258 🔴 (+13%)	1306 🔴 (-8.5%)	30
TTFS	stream	112 (-48%) 💚	1187 🔴 (+9.2%)	1225 🔴 (+12%)	1243 🔴 (+6.9%)	30
TTFS	hook + stream	377 (-69%) 💚	1463 🔴 (+9.8%)	1836 🔴 (+32%) 🔻	2132 🔴 (+36%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	508 (-15%) 💚	1679 (+5.7%)	1705 (-3.1%)	1752 (-6.7%)	10
Fan-out TTLS	Promise.all(100 steps)	1590 (-4.6%)	3569 (+6.8%)	3692 (+4.0%)	4409 (-34%) 💚	10
STSO	1020 steps (inline)	67 (-40%) 💚	133 (-10%)	155 (-11%)	230 (-35%) 💚	1019
WO	1020 steps	131497 (-12%)	131497 (-12%)	131497 (-12%)	131497 (-12%)	1
CRTT	first chunk (pooled)	88 (-9.3%)	120 (-20%) 💚	178 (+3.5%)	355 (+39%) 🔻	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	110 (-25%)	120 (-33%)	185 (-52%)	315 (-58%)	116 (-40%)	10
size sweep (100/s, 160B-12KB)	106 (-17%)	132 (-29%)	268 (-21%)	767 (+62%)	133 (-13%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	117 (-11%)	110 (-31%)	130 (-48%)	224 (-63%)	251 (-24%)	3
replay eve-gpt-5.6-sol-2000t (1x)	115 (-32%)	121 (-29%)	150 (-37%)	335 (-61%)	392 (-36%)	2
replay eve-gpt-5.6-sol-2000t (2x)	120 (-6%)	149 (-38%)	225 (-36%)	457 (-28%)	254 (-11%)	3

📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 148790ms → this run 131259ms (Δ -17531ms, -12%)

 50-100 ms ┃ main 0 this 3 +3
100-150 ms █████████████████████░░┃ main 779 this 872 +93
150-200 ms ███┃█ main 181 this 128 -53
200-250 ms ┃ main 31 this 9 -22
250-300 ms ┃ main 11 this 6 -5
300-350 ms ┃ main 6 this 0 -6
350-400 ms ┃ main 6 this 0 -6
400-450 ms ┃ main 1 this 1 +0
450-500 ms ┃ main 1 this 0 -1
500-550 ms ┃ main 1 this 0 -1
600-650 ms ┃ main 1 this 0 -1
650-700 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)

variant RTT 1ms→5s+ avg p50 p90 p99 n
control ·····▁█▇▁···· 103.2 (-32%) 96 (-24%) 185 (-52%) 315 (-58%) 3000
sweep ······▆█▁▁··· 126.4 (-17%) 107 (-15%) 268 (-21%) 767 (+62%) 3000
gw 1x ·····▁█▅▁···· 95.8 (-30%) 92 (-23%) 130 (-48%) 224 (-63%) 5295
eve 1x ·····▁█▇▁▁··· 103.6 (-35%) 94 (-27%) 150 (-37%) 335 (-61%) 5186
eve 2x ·····▁▅█▂▁··· 128.1 (-32%) 113 (-31%) 225 (-36%) 457 (-28%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▂▃▂▁▅▂▁▃▃ 95–124ms
sweep █▅▃▂▂▄▂▄▂▁ 105–169ms
gw 1x ▃█▃▄▂▂▂▁▂▄ 91–107ms
eve 1x ▃▂▃▃▁█▃▄▂▁ 86–146ms
eve 2x ▂▃▃▁▁▃▆█▅▂ 103–175ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep ▅█▆▃▂▁▂ 123–131ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▅▃▄█▃▅▅▁▃▆ 26–35ms
sweep ▅▆▇▄▇▄▆█▄▁ 41–56ms
gw 1x ▁█▃▃▃▄▆▅▄▅ 23–32ms
eve 1x ▂▃▁▁▂█▅▅▃▃ 19–29ms
eve 2x ▂▄▁▁▁▂█▄▂▆ 19–26ms

ℹ️ Metric definitions & methodology

_{Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.}

_{The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.}

_{The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.}

_{Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.}

_{Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)}

_{Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles}

_{Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600}

_{All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).}

_{Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.}

[github-actions]

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenario	outcome	events	virt	replay	violations
✅ smoke-no-steps	completed	3	0ms	ok	0
✅ smoke-one-step	completed	6	0ms	ok	0
✅ hook-at-step-started	completed	12	0ms	ok	0
✅ hook-at-step-completed	completed	12	0ms	ok	0
✅ hook-at-hook-created	completed	12	0ms	ok	0
✅ deadline-hook-wins	completed	7	1.0h	ok	0
✅ deadline-expires	completed	7	1.0h	ok	0
✅ long-sleep	completed	11	30.0d	ok	0
✅ hook-never-arrives	stalled	3	0ms	skipped	0
✅ step-retries-twice	completed	10	2.0s	ok	0
✅ parallel-steps	completed	9	0ms	ok	0
✅ hook-on-execution-state	completed	12	0ms	ok	0
✅ peek-hook-before-branch	completed	12	0ms	ok	0
✅ peek-hook-after-branch	completed	12	0ms	ok	0
✅ peek-hook-at-registration	completed	12	0ms	ok	0
✅ race-hook-before-probe	completed	12	0ms	ok	0
✅ race-hook-after-probe	completed	12	0ms	ok	0
✅ race-duplicate-delivery	completed	13	0ms	ok	0
✅ attr-hook-before-step	completed	11	0ms	ok	0
✅ attr-hook-after-step	completed	11	0ms	ok	0
✅ attr-from-step-body	completed	13	0ms	ok	0
✅ fork-hook-after-timeout	completed	14	1.0m	ok	0
✅ fork-hook-before-timeout	completed	14	1.0m	ok	0
✅ count-hook-after-timeout	completed	17	1.0m	ok	0
✅ count-hook-before-timeout	completed	20	1.0m	ok	0
✅ stale-read-step-count-fork	completed	20	1.0m	ok	0
✅ stale-read-equal-step-counts	completed	14	1.0m	ok	0
✅ step-vs-step-fork	completed	12	0ms	ok	0
✅ step-vs-step-fork-fenced	completed	12	0ms	ok	0
✅ fence-catches-benign-direction	completed	12	5ms	ok	0
✅ in-flight-before-decision	completed	17	1.0m	ok	0
❌ in-flight-before-decision-counted	completed	17	1.0m	ok	0
✅ in-flight-after-decision	completed	19	2.0m	ok	0
✅ stale-read-step-count-fork-fenced	completed	20	1.0m	ok	0
✅ fork-hook-wins	completed	13	1.0m	ok	0
✅ fork-timeout-wins	completed	13	1.0m	ok	0
✅ unclaimed-payload-under-fork	completed	17	1.0m	ok	0
✅ claimed-payload-under-fork	completed	17	1.0m	ok	0
✅ writers-independent-step-bodies	completed	12	0ms	ok	0
✅ writers-scripted-tempo	completed	12	0ms	ok	0
✅ cancel-mid-step	cancelled	7	0ms	skipped	0

Full trace: world-sim.txt

[vercel]

Additional Suggestion:

Test asserts SPEC_VERSION_CURRENT === SPEC_VERSION_SUPPORTS_COMPRESSION (5) but the PR bumped SPEC_VERSION_CURRENT to SPEC_VERSION_SUPPORTS_SLOT_IDENTITY (6), so the assertion evaluates 6 === 5 and fails.

[github-actions]

Framework	Cold replay	Step reg.	Framework output
hono	304.7 KiB (+103.8 KiB) ⚠️	40.6 KiB (±0)	2.58 MiB (+830.6 KiB)
nextjs-turbopack	305.0 KiB (+98.7 KiB) ⚠️	439 B (±0)	562.1 KiB (-203.4 KiB)

About these numbers

Sizes are gzip; parentheses show the change against main.
Cold replay and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.
⚠️ marks growth past the threshold. Add the allow-bundle-size-growth label to accept it.

770e721 · run

[NathanColosimo]

High Level

Change build output to be separate workflow bundles and load only the needed code in the runtime.
There's very few changes to the runtime code, most of the changes are in the builder code.
Almost all of the changes are tests

Summary

• emit one VM bundle per workflow source; workflows in the same source share one bundle
• keep one flow route with a static workflow-ID → dynamic-import loader map
• overlap the selected bundle load with world initialization and event setup
• skip bundle loading for queue-only step deliveries
• retain compiled scripts for every immutable production source bundle
• keep watch loaders cache-safe and derive loader IDs from the same transform that emitted each bundle
• encode generated VM modules as opaque artifacts so framework plugins cannot rewrite inert source

Result

For the Next/Turbopack benchmark app's 97_bench source, measured against the monolithic bundle at the base of this PR:

• generated flow route: 1,327,514 B → 18,264 B (-98.6%)
• selected benchmark VM code: 1,307,595 B → 133,859 B (-89.8%)
• cold compile p50: 10.405 ms → 1.089 ms (-89.5%)
• fresh-context evaluation p50: 7.484 ms → 0.356 ms (-95.2%)
• one-time base64 artifact decode p50/p90: 0.037/0.044 ms

The compile/evaluation microbenchmarks use 60 uncached vm.Script compilations and 80 evaluations in fresh VM contexts on Node. The deployment benchmark and APM traces remain the end-to-end signal.

The correctness-preserving split deliberately includes any source that defines both workflows and custom serializers in every other source bundle. The workbench's unusually large 99_e2e.ts is such a hybrid, so the selected benchmark bundle is ~134 KB rather than the ~28 KB it would be without that shared registration. The next structural improvement is to extract serializer registration into a dedicated small VM prelude; omitting it would make replay deserialization incorrect.

The 19 split bundles total 5.28 MB decoded versus the old 1.31 MB monolith because shared sandbox dependencies and the hybrid serializer source repeat. Opaque base64 transport is 7.04 MB on disk before compression (1.33× decoded; gzip is close to the original text). This is a build/storage tradeoff: a replay imports, decodes, compiles, and evaluates only its selected source bundle, and the route caches that decoded promise.

Deployment benchmark

The final benchmark run completed successfully on deployment dpl_CHJqdRSi52QvbKm1KzXhEs4Aa1B2; the sticky comparison reports:

• inline STSO p75 191 → 169 ms (-12%), p90 229 → 191 ms (-17%), p99 580 → 268 ms (-54%)
• whole-run overhead for 1,020 steps 195,405 → 167,186 ms (-14%)
• an earlier run of the same runtime code measured STSO p75/p90/p99 at 155/172/264 ms and whole-run overhead at 153,275 ms, so the improvement survives expected cold-start and infrastructure variance
• TTFS p75 remains 1.12–1.50 s across scenarios, showing that dispatch, route cold start, and workflow-server work still dominate end-to-end startup

A final cold compile-miss trace has a 22.52 ms workflow.run; the surrounding fresh-replay path shows 6.88 ms bundle load (overlapped with 2.37 ms world init), followed by 3.00 ms context creation, 4.00 ms compile, 2.63 ms evaluate, 1.53 ms input hydrate, and 8.82 ms replay execute. Its 146 ms payload-preparation wall span begins before the replay and overlaps event loading and execution rather than representing 146 ms of blocking CPU work.

After the bundle and script caches are warm, the final sequential benchmark's first fresh VM replay trace is 6.88 ms: context creation 2.00 ms, compile hit 0.14 ms, evaluation 1.94 ms, input hydration 0.12 ms, and replay execution 2.41 ms. The comparable main trace is 186.10 ms (-96.3% for this representative warm-cache replay). Subsequent retained invocations in the final trace are generally 1–2.4 ms.

Validation

• 295 focused tests across builders, runtime, tracing, payload cache, and stream recovery
• complete builders suite: 226/226
• affected package typechecks and builds
• Next/Turbopack, Vite, and Nuxt production builds
• exact Next HMR regression plus same-size/same-mtime workflow-ID rename regression
• all eight stable/canary × Webpack/Turbopack × Node/QuickJS Next dev E2E lanes green
• Linux and Windows unit lanes plus Node/QuickJS Windows Next E2E lanes green
• standalone/BOPA import smoke test
• final benchmark workflow green
• final Codex and Claude autoreviews clean

[changeset-bot]

🦋 Changeset detected

Latest commit: 770e721

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 16 packages

Name	Type
@workflow/builders	Patch
@workflow/core	Patch
@workflow/next	Patch
@workflow/world-testing	Patch
workflow	Patch
@workflow/astro	Patch
@workflow/cli	Patch
@workflow/nest	Patch
@workflow/nitro	Patch
@workflow/rollup	Patch
@workflow/sveltekit	Patch
@workflow/vite	Patch
@workflow/vitest	Patch
@workflow/web-shared	Patch
@workflow/web	Patch
@workflow/nuxt	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-express-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nestjs-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-python-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-docs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-swc-playground	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-tarballs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-web	Ready	Preview, v0	Sep 1, 2026 3:32am UTC

[github-actions]

🧪 E2E Test Results

❌ Some tests failed

❌ Failed E2E Tests

▲ Vercel Production (8 failed)

python-node (8 failed):

• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3 | 🔍 observability
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M | 🔍 observability
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23 | 🔍 observability
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG | 🔍 observability
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115 | 🔍 observability
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP | 🔍 observability
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY | 🔍 observability
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF | 🔍 observability

🌐 Cross-language Conformance (9 failed)

python (9 failed):

• deploymentId: 'latest' is a no-op in non-Vercel worlds | wrun_01M1DGN5VWSB6FR072SCMVKD93
• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

• cancelRun via CLI - cancelling a running workflow (nextjs-turbopack)
• cancelRun via CLI - cancelling a running workflow (nuxt)
• promiseAllWorkflow (nest)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

31 infra events

• cold-start-warmup · suite warmup (python) · at 03:32:38Z · abandoned wrun_41M1DG9NKP0GT49DFD2FT24FQP · (+7 more)
• run-pickup-stall · nullByteWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA4A0GWNBVG1J3PNW2BM
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F7
• run-pickup-stall · sleepingWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F6
• run-pickup-stall · promiseAllWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA410GG3J9DQS5J8V5RX
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDACG0GYN2PV31FE1722G
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:33:26Z · abandoned wrun_41M1DGE9JE0GYK5Q2XK9JWVR87
• run-pickup-stall · sleepingWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5DP0GZT0NGX0185G3X8
• run-pickup-stall · promiseAllWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5BE0GMZ14CQP9SC1Y7K
• run-pickup-stall · nullByteWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5AD0GMMXJ2AJ6AZN3XB
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:33:55Z · abandoned wrun_41M1DGF5BM0GJKW6R6A11PP5G5
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:34:45Z · abandoned wrun_41M1DGF8SE0GTQDHMN40HTGC3X
• cold-start-warmup · suite warmup (python) · at 03:35:40Z · abandoned wrun_01M1DGF7EGPY04C8PJ02V841CN · (+7 more)
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJV4MD7KSP260W7ZSBC
• run-pickup-stall · sleepingWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK0G2QC0ZT6CH9PZYC0
• run-pickup-stall · promiseAllWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJW8G1Y4FFC31VK77KY
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK38F43AA3AECZXYP9H
• run-pickup-stall · nullByteWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK5160SZJ5HM2H6FMB9
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:36:26Z · abandoned wrun_41M1DGH6TB0GZCNX8E7RCMJJ6N
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:36:26Z · abandoned wrun_41M1DGJ0P00GPSJ9R5XE1S1WHX
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6HAQR493S719104TSK
• run-pickup-stall · sleepingWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6NBAJ7B203HQDK5N2N
• run-pickup-stall · promiseAllWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6KR8J5R749MV6Q056B
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6SZKHYHZ67N6HHVCTV
• run-pickup-stall · nullByteWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ70B63HA9BG628J3E8C
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHT45WP6FSWB7FGZRVDY
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTEREXVJRZ204H80RN5
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTKKDTFSEZV6GQKNE3R
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF60XCRG4QP1WMRPCM2K
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF5YJJWK6M7RQPW03HKJ
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:38:55Z · abandoned wrun_01M1DGRCDP4DNKXYYTR30VTRPT

E2E Test Summary

Summary

	Passed	Failed	Skipped	Total
❌ ▲ Vercel Production	3570	8	742	4320
✅ 💻 Local Development	3922	0	558	4480
✅ 📦 Local Production	3922	0	558	4480
✅ 🐘 Local Postgres	3922	0	558	4480
✅ 🪟 Windows	320	0	0	320
❌ 🌐 Cross-language Conformance	0	9	132	141
✅ vercel-http-transport	817	0	143	960
✅ vercel-multi-region	27	0	0	27
✅ vercel-ws-transport	553	0	87	640
Total	17053	17	2778	19848

Details by Category

❌ ▲ Vercel Production

App	Passed	Failed	Skipped
✅ astro-node	132	0	28
✅ astro-quickjs	132	0	28
✅ example-node	132	0	28
✅ example-quickjs	132	0	28
✅ express-node	132	0	28
✅ express-quickjs	132	0	28
✅ fastify-node	132	0	28
✅ fastify-quickjs	132	0	28
✅ hono-node	132	0	28
✅ hono-quickjs	132	0	28
✅ nest-node	132	0	28
✅ nest-quickjs	132	0	28
✅ nextjs-turbopack-node	157	0	3
✅ nextjs-turbopack-quickjs	157	0	3
✅ nextjs-webpack-node	157	0	3
✅ nextjs-webpack-quickjs	157	0	3
✅ nitro-node	132	0	28
✅ nitro-quickjs	132	0	28
✅ nuxt-node	132	0	28
✅ nuxt-quickjs	132	0	28
❌ python-node	0	8	152
✅ sveltekit-node	151	0	9
✅ sveltekit-quickjs	151	0	9
✅ tanstack-start-node	132	0	28
✅ tanstack-start-quickjs	132	0	28
✅ vite-node	132	0	28
✅ vite-quickjs	132	0	28

✅ 💻 Local Development

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 📦 Local Production

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🐘 Local Postgres

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	160	0	0
✅ nextjs-turbopack-quickjs	160	0	0

❌ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
❌ python	0	9	132

✅ vercel-http-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ hono	132	0	28
✅ nextjs-turbopack	157	0	3
✅ nitro	132	0	28
✅ vite	132	0	28

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ nextjs-turbopack	157	0	3
✅ vite	132	0	28

---

📋 View full workflow run

[github-actions]

📊 Workflow Benchmarks

commit 770e721 · Tue, 01 Sep 2026 03:46:20 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	1102 (+502%) 🔻	1224 🔴 (+15%) 🔻	1258 🔴 (+13%)	1306 🔴 (-8.5%)	30
TTFS	stream	112 (-48%) 💚	1187 🔴 (+9.2%)	1225 🔴 (+12%)	1243 🔴 (+6.9%)	30
TTFS	hook + stream	377 (-69%) 💚	1463 🔴 (+9.8%)	1836 🔴 (+32%) 🔻	2132 🔴 (+36%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	508 (-15%) 💚	1679 (+5.7%)	1705 (-3.1%)	1752 (-6.7%)	10
Fan-out TTLS	Promise.all(100 steps)	1590 (-4.6%)	3569 (+6.8%)	3692 (+4.0%)	4409 (-34%) 💚	10
STSO	1020 steps (inline)	67 (-40%) 💚	133 (-10%)	155 (-11%)	230 (-35%) 💚	1019
WO	1020 steps	131497 (-12%)	131497 (-12%)	131497 (-12%)	131497 (-12%)	1
CRTT	first chunk (pooled)	88 (-9.3%)	120 (-20%) 💚	178 (+3.5%)	355 (+39%) 🔻	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	110 (-25%)	120 (-33%)	185 (-52%)	315 (-58%)	116 (-40%)	10
size sweep (100/s, 160B-12KB)	106 (-17%)	132 (-29%)	268 (-21%)	767 (+62%)	133 (-13%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	117 (-11%)	110 (-31%)	130 (-48%)	224 (-63%)	251 (-24%)	3
replay eve-gpt-5.6-sol-2000t (1x)	115 (-32%)	121 (-29%)	150 (-37%)	335 (-61%)	392 (-36%)	2
replay eve-gpt-5.6-sol-2000t (2x)	120 (-6%)	149 (-38%)	225 (-36%)	457 (-28%)	254 (-11%)	3

📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 148790ms → this run 131259ms (Δ -17531ms, -12%)

 50-100 ms ┃ main 0 this 3 +3
100-150 ms █████████████████████░░┃ main 779 this 872 +93
150-200 ms ███┃█ main 181 this 128 -53
200-250 ms ┃ main 31 this 9 -22
250-300 ms ┃ main 11 this 6 -5
300-350 ms ┃ main 6 this 0 -6
350-400 ms ┃ main 6 this 0 -6
400-450 ms ┃ main 1 this 1 +0
450-500 ms ┃ main 1 this 0 -1
500-550 ms ┃ main 1 this 0 -1
600-650 ms ┃ main 1 this 0 -1
650-700 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)

variant RTT 1ms→5s+ avg p50 p90 p99 n
control ·····▁█▇▁···· 103.2 (-32%) 96 (-24%) 185 (-52%) 315 (-58%) 3000
sweep ······▆█▁▁··· 126.4 (-17%) 107 (-15%) 268 (-21%) 767 (+62%) 3000
gw 1x ·····▁█▅▁···· 95.8 (-30%) 92 (-23%) 130 (-48%) 224 (-63%) 5295
eve 1x ·····▁█▇▁▁··· 103.6 (-35%) 94 (-27%) 150 (-37%) 335 (-61%) 5186
eve 2x ·····▁▅█▂▁··· 128.1 (-32%) 113 (-31%) 225 (-36%) 457 (-28%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▂▃▂▁▅▂▁▃▃ 95–124ms
sweep █▅▃▂▂▄▂▄▂▁ 105–169ms
gw 1x ▃█▃▄▂▂▂▁▂▄ 91–107ms
eve 1x ▃▂▃▃▁█▃▄▂▁ 86–146ms
eve 2x ▂▃▃▁▁▃▆█▅▂ 103–175ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep ▅█▆▃▂▁▂ 123–131ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▅▃▄█▃▅▅▁▃▆ 26–35ms
sweep ▅▆▇▄▇▄▆█▄▁ 41–56ms
gw 1x ▁█▃▃▃▄▆▅▄▅ 23–32ms
eve 1x ▂▃▁▁▂█▅▅▃▃ 19–29ms
eve 2x ▂▄▁▁▁▂█▄▂▆ 19–26ms

ℹ️ Metric definitions & methodology

_{Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.}

_{The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.}

_{The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.}

_{Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.}

_{Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)}

_{Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles}

_{Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600}

_{All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).}

_{Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.}

[github-actions]

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenario	outcome	events	virt	replay	violations
✅ smoke-no-steps	completed	3	0ms	ok	0
✅ smoke-one-step	completed	6	0ms	ok	0
✅ hook-at-step-started	completed	12	0ms	ok	0
✅ hook-at-step-completed	completed	12	0ms	ok	0
✅ hook-at-hook-created	completed	12	0ms	ok	0
✅ deadline-hook-wins	completed	7	1.0h	ok	0
✅ deadline-expires	completed	7	1.0h	ok	0
✅ long-sleep	completed	11	30.0d	ok	0
✅ hook-never-arrives	stalled	3	0ms	skipped	0
✅ step-retries-twice	completed	10	2.0s	ok	0
✅ parallel-steps	completed	9	0ms	ok	0
✅ hook-on-execution-state	completed	12	0ms	ok	0
✅ peek-hook-before-branch	completed	12	0ms	ok	0
✅ peek-hook-after-branch	completed	12	0ms	ok	0
✅ peek-hook-at-registration	completed	12	0ms	ok	0
✅ race-hook-before-probe	completed	12	0ms	ok	0
✅ race-hook-after-probe	completed	12	0ms	ok	0
✅ race-duplicate-delivery	completed	13	0ms	ok	0
✅ attr-hook-before-step	completed	11	0ms	ok	0
✅ attr-hook-after-step	completed	11	0ms	ok	0
✅ attr-from-step-body	completed	13	0ms	ok	0
✅ fork-hook-after-timeout	completed	14	1.0m	ok	0
✅ fork-hook-before-timeout	completed	14	1.0m	ok	0
✅ count-hook-after-timeout	completed	17	1.0m	ok	0
✅ count-hook-before-timeout	completed	20	1.0m	ok	0
✅ stale-read-step-count-fork	completed	20	1.0m	ok	0
✅ stale-read-equal-step-counts	completed	14	1.0m	ok	0
✅ step-vs-step-fork	completed	12	0ms	ok	0
✅ step-vs-step-fork-fenced	completed	12	0ms	ok	0
✅ fence-catches-benign-direction	completed	12	5ms	ok	0
✅ in-flight-before-decision	completed	17	1.0m	ok	0
❌ in-flight-before-decision-counted	completed	17	1.0m	ok	0
✅ in-flight-after-decision	completed	19	2.0m	ok	0
✅ stale-read-step-count-fork-fenced	completed	20	1.0m	ok	0
✅ fork-hook-wins	completed	13	1.0m	ok	0
✅ fork-timeout-wins	completed	13	1.0m	ok	0
✅ unclaimed-payload-under-fork	completed	17	1.0m	ok	0
✅ claimed-payload-under-fork	completed	17	1.0m	ok	0
✅ writers-independent-step-bodies	completed	12	0ms	ok	0
✅ writers-scripted-tempo	completed	12	0ms	ok	0
✅ cancel-mid-step	cancelled	7	0ms	skipped	0

Full trace: world-sim.txt

[vercel]

Additional Suggestion:

Test asserts SPEC_VERSION_CURRENT === SPEC_VERSION_SUPPORTS_COMPRESSION (5) but the PR bumped SPEC_VERSION_CURRENT to SPEC_VERSION_SUPPORTS_SLOT_IDENTITY (6), so the assertion evaluates 6 === 5 and fails.

[github-actions]

Framework	Cold replay	Step reg.	Framework output
hono	304.7 KiB (+103.8 KiB) ⚠️	40.6 KiB (±0)	2.58 MiB (+830.6 KiB)
nextjs-turbopack	305.0 KiB (+98.7 KiB) ⚠️	439 B (±0)	562.1 KiB (-203.4 KiB)

About these numbers

Sizes are gzip; parentheses show the change against main.
Cold replay and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.
⚠️ marks growth past the threshold. Add the allow-bundle-size-growth label to accept it.

770e721 · run

[NathanColosimo]

High Level

Change build output to be separate workflow bundles and load only the needed code in the runtime.
There's very few changes to the runtime code, most of the changes are in the builder code.
Almost all of the changes are tests

Summary

• emit one VM bundle per workflow source; workflows in the same source share one bundle
• keep one flow route with a static workflow-ID → dynamic-import loader map
• overlap the selected bundle load with world initialization and event setup
• skip bundle loading for queue-only step deliveries
• retain compiled scripts for every immutable production source bundle
• keep watch loaders cache-safe and derive loader IDs from the same transform that emitted each bundle
• encode generated VM modules as opaque artifacts so framework plugins cannot rewrite inert source

Result

For the Next/Turbopack benchmark app's 97_bench source, measured against the monolithic bundle at the base of this PR:

• generated flow route: 1,327,514 B → 18,264 B (-98.6%)
• selected benchmark VM code: 1,307,595 B → 133,859 B (-89.8%)
• cold compile p50: 10.405 ms → 1.089 ms (-89.5%)
• fresh-context evaluation p50: 7.484 ms → 0.356 ms (-95.2%)
• one-time base64 artifact decode p50/p90: 0.037/0.044 ms

The compile/evaluation microbenchmarks use 60 uncached vm.Script compilations and 80 evaluations in fresh VM contexts on Node. The deployment benchmark and APM traces remain the end-to-end signal.

The correctness-preserving split deliberately includes any source that defines both workflows and custom serializers in every other source bundle. The workbench's unusually large 99_e2e.ts is such a hybrid, so the selected benchmark bundle is ~134 KB rather than the ~28 KB it would be without that shared registration. The next structural improvement is to extract serializer registration into a dedicated small VM prelude; omitting it would make replay deserialization incorrect.

The 19 split bundles total 5.28 MB decoded versus the old 1.31 MB monolith because shared sandbox dependencies and the hybrid serializer source repeat. Opaque base64 transport is 7.04 MB on disk before compression (1.33× decoded; gzip is close to the original text). This is a build/storage tradeoff: a replay imports, decodes, compiles, and evaluates only its selected source bundle, and the route caches that decoded promise.

Deployment benchmark

The final benchmark run completed successfully on deployment dpl_CHJqdRSi52QvbKm1KzXhEs4Aa1B2; the sticky comparison reports:

• inline STSO p75 191 → 169 ms (-12%), p90 229 → 191 ms (-17%), p99 580 → 268 ms (-54%)
• whole-run overhead for 1,020 steps 195,405 → 167,186 ms (-14%)
• an earlier run of the same runtime code measured STSO p75/p90/p99 at 155/172/264 ms and whole-run overhead at 153,275 ms, so the improvement survives expected cold-start and infrastructure variance
• TTFS p75 remains 1.12–1.50 s across scenarios, showing that dispatch, route cold start, and workflow-server work still dominate end-to-end startup

A final cold compile-miss trace has a 22.52 ms workflow.run; the surrounding fresh-replay path shows 6.88 ms bundle load (overlapped with 2.37 ms world init), followed by 3.00 ms context creation, 4.00 ms compile, 2.63 ms evaluate, 1.53 ms input hydrate, and 8.82 ms replay execute. Its 146 ms payload-preparation wall span begins before the replay and overlaps event loading and execution rather than representing 146 ms of blocking CPU work.

After the bundle and script caches are warm, the final sequential benchmark's first fresh VM replay trace is 6.88 ms: context creation 2.00 ms, compile hit 0.14 ms, evaluation 1.94 ms, input hydration 0.12 ms, and replay execution 2.41 ms. The comparable main trace is 186.10 ms (-96.3% for this representative warm-cache replay). Subsequent retained invocations in the final trace are generally 1–2.4 ms.

Validation

• 295 focused tests across builders, runtime, tracing, payload cache, and stream recovery
• complete builders suite: 226/226
• affected package typechecks and builds
• Next/Turbopack, Vite, and Nuxt production builds
• exact Next HMR regression plus same-size/same-mtime workflow-ID rename regression
• all eight stable/canary × Webpack/Turbopack × Node/QuickJS Next dev E2E lanes green
• Linux and Windows unit lanes plus Node/QuickJS Windows Next E2E lanes green
• standalone/BOPA import smoke test
• final benchmark workflow green
• final Codex and Claude autoreviews clean

[changeset-bot]

🦋 Changeset detected

Latest commit: 770e721

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 16 packages

Name	Type
@workflow/builders	Patch
@workflow/core	Patch
@workflow/next	Patch
@workflow/world-testing	Patch
workflow	Patch
@workflow/astro	Patch
@workflow/cli	Patch
@workflow/nest	Patch
@workflow/nitro	Patch
@workflow/rollup	Patch
@workflow/sveltekit	Patch
@workflow/vite	Patch
@workflow/vitest	Patch
@workflow/web-shared	Patch
@workflow/web	Patch
@workflow/nuxt	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-express-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nestjs-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-python-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-docs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-swc-playground	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-tarballs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-web	Ready	Preview, v0	Sep 1, 2026 3:32am UTC

[github-actions]

🧪 E2E Test Results

❌ Some tests failed

❌ Failed E2E Tests

▲ Vercel Production (8 failed)

python-node (8 failed):

• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3 | 🔍 observability
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M | 🔍 observability
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23 | 🔍 observability
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG | 🔍 observability
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115 | 🔍 observability
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP | 🔍 observability
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY | 🔍 observability
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF | 🔍 observability

🌐 Cross-language Conformance (9 failed)

python (9 failed):

• deploymentId: 'latest' is a no-op in non-Vercel worlds | wrun_01M1DGN5VWSB6FR072SCMVKD93
• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

• cancelRun via CLI - cancelling a running workflow (nextjs-turbopack)
• cancelRun via CLI - cancelling a running workflow (nuxt)
• promiseAllWorkflow (nest)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

31 infra events

• cold-start-warmup · suite warmup (python) · at 03:32:38Z · abandoned wrun_41M1DG9NKP0GT49DFD2FT24FQP · (+7 more)
• run-pickup-stall · nullByteWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA4A0GWNBVG1J3PNW2BM
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F7
• run-pickup-stall · sleepingWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F6
• run-pickup-stall · promiseAllWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA410GG3J9DQS5J8V5RX
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDACG0GYN2PV31FE1722G
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:33:26Z · abandoned wrun_41M1DGE9JE0GYK5Q2XK9JWVR87
• run-pickup-stall · sleepingWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5DP0GZT0NGX0185G3X8
• run-pickup-stall · promiseAllWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5BE0GMZ14CQP9SC1Y7K
• run-pickup-stall · nullByteWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5AD0GMMXJ2AJ6AZN3XB
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:33:55Z · abandoned wrun_41M1DGF5BM0GJKW6R6A11PP5G5
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:34:45Z · abandoned wrun_41M1DGF8SE0GTQDHMN40HTGC3X
• cold-start-warmup · suite warmup (python) · at 03:35:40Z · abandoned wrun_01M1DGF7EGPY04C8PJ02V841CN · (+7 more)
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJV4MD7KSP260W7ZSBC
• run-pickup-stall · sleepingWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK0G2QC0ZT6CH9PZYC0
• run-pickup-stall · promiseAllWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJW8G1Y4FFC31VK77KY
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK38F43AA3AECZXYP9H
• run-pickup-stall · nullByteWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK5160SZJ5HM2H6FMB9
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:36:26Z · abandoned wrun_41M1DGH6TB0GZCNX8E7RCMJJ6N
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:36:26Z · abandoned wrun_41M1DGJ0P00GPSJ9R5XE1S1WHX
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6HAQR493S719104TSK
• run-pickup-stall · sleepingWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6NBAJ7B203HQDK5N2N
• run-pickup-stall · promiseAllWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6KR8J5R749MV6Q056B
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6SZKHYHZ67N6HHVCTV
• run-pickup-stall · nullByteWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ70B63HA9BG628J3E8C
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHT45WP6FSWB7FGZRVDY
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTEREXVJRZ204H80RN5
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTKKDTFSEZV6GQKNE3R
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF60XCRG4QP1WMRPCM2K
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF5YJJWK6M7RQPW03HKJ
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:38:55Z · abandoned wrun_01M1DGRCDP4DNKXYYTR30VTRPT

E2E Test Summary

Summary

	Passed	Failed	Skipped	Total
❌ ▲ Vercel Production	3570	8	742	4320
✅ 💻 Local Development	3922	0	558	4480
✅ 📦 Local Production	3922	0	558	4480
✅ 🐘 Local Postgres	3922	0	558	4480
✅ 🪟 Windows	320	0	0	320
❌ 🌐 Cross-language Conformance	0	9	132	141
✅ vercel-http-transport	817	0	143	960
✅ vercel-multi-region	27	0	0	27
✅ vercel-ws-transport	553	0	87	640
Total	17053	17	2778	19848

Details by Category

❌ ▲ Vercel Production

App	Passed	Failed	Skipped
✅ astro-node	132	0	28
✅ astro-quickjs	132	0	28
✅ example-node	132	0	28
✅ example-quickjs	132	0	28
✅ express-node	132	0	28
✅ express-quickjs	132	0	28
✅ fastify-node	132	0	28
✅ fastify-quickjs	132	0	28
✅ hono-node	132	0	28
✅ hono-quickjs	132	0	28
✅ nest-node	132	0	28
✅ nest-quickjs	132	0	28
✅ nextjs-turbopack-node	157	0	3
✅ nextjs-turbopack-quickjs	157	0	3
✅ nextjs-webpack-node	157	0	3
✅ nextjs-webpack-quickjs	157	0	3
✅ nitro-node	132	0	28
✅ nitro-quickjs	132	0	28
✅ nuxt-node	132	0	28
✅ nuxt-quickjs	132	0	28
❌ python-node	0	8	152
✅ sveltekit-node	151	0	9
✅ sveltekit-quickjs	151	0	9
✅ tanstack-start-node	132	0	28
✅ tanstack-start-quickjs	132	0	28
✅ vite-node	132	0	28
✅ vite-quickjs	132	0	28

✅ 💻 Local Development

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 📦 Local Production

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🐘 Local Postgres

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	160	0	0
✅ nextjs-turbopack-quickjs	160	0	0

❌ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
❌ python	0	9	132

✅ vercel-http-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ hono	132	0	28
✅ nextjs-turbopack	157	0	3
✅ nitro	132	0	28
✅ vite	132	0	28

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ nextjs-turbopack	157	0	3
✅ vite	132	0	28

---

📋 View full workflow run

[github-actions]

📊 Workflow Benchmarks

commit 770e721 · Tue, 01 Sep 2026 03:46:20 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	1102 (+502%) 🔻	1224 🔴 (+15%) 🔻	1258 🔴 (+13%)	1306 🔴 (-8.5%)	30
TTFS	stream	112 (-48%) 💚	1187 🔴 (+9.2%)	1225 🔴 (+12%)	1243 🔴 (+6.9%)	30
TTFS	hook + stream	377 (-69%) 💚	1463 🔴 (+9.8%)	1836 🔴 (+32%) 🔻	2132 🔴 (+36%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	508 (-15%) 💚	1679 (+5.7%)	1705 (-3.1%)	1752 (-6.7%)	10
Fan-out TTLS	Promise.all(100 steps)	1590 (-4.6%)	3569 (+6.8%)	3692 (+4.0%)	4409 (-34%) 💚	10
STSO	1020 steps (inline)	67 (-40%) 💚	133 (-10%)	155 (-11%)	230 (-35%) 💚	1019
WO	1020 steps	131497 (-12%)	131497 (-12%)	131497 (-12%)	131497 (-12%)	1
CRTT	first chunk (pooled)	88 (-9.3%)	120 (-20%) 💚	178 (+3.5%)	355 (+39%) 🔻	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	110 (-25%)	120 (-33%)	185 (-52%)	315 (-58%)	116 (-40%)	10
size sweep (100/s, 160B-12KB)	106 (-17%)	132 (-29%)	268 (-21%)	767 (+62%)	133 (-13%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	117 (-11%)	110 (-31%)	130 (-48%)	224 (-63%)	251 (-24%)	3
replay eve-gpt-5.6-sol-2000t (1x)	115 (-32%)	121 (-29%)	150 (-37%)	335 (-61%)	392 (-36%)	2
replay eve-gpt-5.6-sol-2000t (2x)	120 (-6%)	149 (-38%)	225 (-36%)	457 (-28%)	254 (-11%)	3

📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 148790ms → this run 131259ms (Δ -17531ms, -12%)

 50-100 ms ┃ main 0 this 3 +3
100-150 ms █████████████████████░░┃ main 779 this 872 +93
150-200 ms ███┃█ main 181 this 128 -53
200-250 ms ┃ main 31 this 9 -22
250-300 ms ┃ main 11 this 6 -5
300-350 ms ┃ main 6 this 0 -6
350-400 ms ┃ main 6 this 0 -6
400-450 ms ┃ main 1 this 1 +0
450-500 ms ┃ main 1 this 0 -1
500-550 ms ┃ main 1 this 0 -1
600-650 ms ┃ main 1 this 0 -1
650-700 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)

variant RTT 1ms→5s+ avg p50 p90 p99 n
control ·····▁█▇▁···· 103.2 (-32%) 96 (-24%) 185 (-52%) 315 (-58%) 3000
sweep ······▆█▁▁··· 126.4 (-17%) 107 (-15%) 268 (-21%) 767 (+62%) 3000
gw 1x ·····▁█▅▁···· 95.8 (-30%) 92 (-23%) 130 (-48%) 224 (-63%) 5295
eve 1x ·····▁█▇▁▁··· 103.6 (-35%) 94 (-27%) 150 (-37%) 335 (-61%) 5186
eve 2x ·····▁▅█▂▁··· 128.1 (-32%) 113 (-31%) 225 (-36%) 457 (-28%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▂▃▂▁▅▂▁▃▃ 95–124ms
sweep █▅▃▂▂▄▂▄▂▁ 105–169ms
gw 1x ▃█▃▄▂▂▂▁▂▄ 91–107ms
eve 1x ▃▂▃▃▁█▃▄▂▁ 86–146ms
eve 2x ▂▃▃▁▁▃▆█▅▂ 103–175ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep ▅█▆▃▂▁▂ 123–131ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▅▃▄█▃▅▅▁▃▆ 26–35ms
sweep ▅▆▇▄▇▄▆█▄▁ 41–56ms
gw 1x ▁█▃▃▃▄▆▅▄▅ 23–32ms
eve 1x ▂▃▁▁▂█▅▅▃▃ 19–29ms
eve 2x ▂▄▁▁▁▂█▄▂▆ 19–26ms

ℹ️ Metric definitions & methodology

_{Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.}

_{The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.}

_{The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.}

_{Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.}

_{Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)}

_{Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles}

_{Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600}

_{All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).}

_{Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.}

[github-actions]

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenario	outcome	events	virt	replay	violations
✅ smoke-no-steps	completed	3	0ms	ok	0
✅ smoke-one-step	completed	6	0ms	ok	0
✅ hook-at-step-started	completed	12	0ms	ok	0
✅ hook-at-step-completed	completed	12	0ms	ok	0
✅ hook-at-hook-created	completed	12	0ms	ok	0
✅ deadline-hook-wins	completed	7	1.0h	ok	0
✅ deadline-expires	completed	7	1.0h	ok	0
✅ long-sleep	completed	11	30.0d	ok	0
✅ hook-never-arrives	stalled	3	0ms	skipped	0
✅ step-retries-twice	completed	10	2.0s	ok	0
✅ parallel-steps	completed	9	0ms	ok	0
✅ hook-on-execution-state	completed	12	0ms	ok	0
✅ peek-hook-before-branch	completed	12	0ms	ok	0
✅ peek-hook-after-branch	completed	12	0ms	ok	0
✅ peek-hook-at-registration	completed	12	0ms	ok	0
✅ race-hook-before-probe	completed	12	0ms	ok	0
✅ race-hook-after-probe	completed	12	0ms	ok	0
✅ race-duplicate-delivery	completed	13	0ms	ok	0
✅ attr-hook-before-step	completed	11	0ms	ok	0
✅ attr-hook-after-step	completed	11	0ms	ok	0
✅ attr-from-step-body	completed	13	0ms	ok	0
✅ fork-hook-after-timeout	completed	14	1.0m	ok	0
✅ fork-hook-before-timeout	completed	14	1.0m	ok	0
✅ count-hook-after-timeout	completed	17	1.0m	ok	0
✅ count-hook-before-timeout	completed	20	1.0m	ok	0
✅ stale-read-step-count-fork	completed	20	1.0m	ok	0
✅ stale-read-equal-step-counts	completed	14	1.0m	ok	0
✅ step-vs-step-fork	completed	12	0ms	ok	0
✅ step-vs-step-fork-fenced	completed	12	0ms	ok	0
✅ fence-catches-benign-direction	completed	12	5ms	ok	0
✅ in-flight-before-decision	completed	17	1.0m	ok	0
❌ in-flight-before-decision-counted	completed	17	1.0m	ok	0
✅ in-flight-after-decision	completed	19	2.0m	ok	0
✅ stale-read-step-count-fork-fenced	completed	20	1.0m	ok	0
✅ fork-hook-wins	completed	13	1.0m	ok	0
✅ fork-timeout-wins	completed	13	1.0m	ok	0
✅ unclaimed-payload-under-fork	completed	17	1.0m	ok	0
✅ claimed-payload-under-fork	completed	17	1.0m	ok	0
✅ writers-independent-step-bodies	completed	12	0ms	ok	0
✅ writers-scripted-tempo	completed	12	0ms	ok	0
✅ cancel-mid-step	cancelled	7	0ms	skipped	0

Full trace: world-sim.txt

[vercel]

Additional Suggestion:

Test asserts SPEC_VERSION_CURRENT === SPEC_VERSION_SUPPORTS_COMPRESSION (5) but the PR bumped SPEC_VERSION_CURRENT to SPEC_VERSION_SUPPORTS_SLOT_IDENTITY (6), so the assertion evaluates 6 === 5 and fails.

[github-actions]

Framework	Cold replay	Step reg.	Framework output
hono	304.7 KiB (+103.8 KiB) ⚠️	40.6 KiB (±0)	2.58 MiB (+830.6 KiB)
nextjs-turbopack	305.0 KiB (+98.7 KiB) ⚠️	439 B (±0)	562.1 KiB (-203.4 KiB)

About these numbers

Sizes are gzip; parentheses show the change against main.
Cold replay and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.
⚠️ marks growth past the threshold. Add the allow-bundle-size-growth label to accept it.

770e721 · run

[NathanColosimo]

High Level

Change build output to be separate workflow bundles and load only the needed code in the runtime.
There's very few changes to the runtime code, most of the changes are in the builder code.
Almost all of the changes are tests

Summary

• emit one VM bundle per workflow source; workflows in the same source share one bundle
• keep one flow route with a static workflow-ID → dynamic-import loader map
• overlap the selected bundle load with world initialization and event setup
• skip bundle loading for queue-only step deliveries
• retain compiled scripts for every immutable production source bundle
• keep watch loaders cache-safe and derive loader IDs from the same transform that emitted each bundle
• encode generated VM modules as opaque artifacts so framework plugins cannot rewrite inert source

Result

For the Next/Turbopack benchmark app's 97_bench source, measured against the monolithic bundle at the base of this PR:

• generated flow route: 1,327,514 B → 18,264 B (-98.6%)
• selected benchmark VM code: 1,307,595 B → 133,859 B (-89.8%)
• cold compile p50: 10.405 ms → 1.089 ms (-89.5%)
• fresh-context evaluation p50: 7.484 ms → 0.356 ms (-95.2%)
• one-time base64 artifact decode p50/p90: 0.037/0.044 ms

The compile/evaluation microbenchmarks use 60 uncached vm.Script compilations and 80 evaluations in fresh VM contexts on Node. The deployment benchmark and APM traces remain the end-to-end signal.

The correctness-preserving split deliberately includes any source that defines both workflows and custom serializers in every other source bundle. The workbench's unusually large 99_e2e.ts is such a hybrid, so the selected benchmark bundle is ~134 KB rather than the ~28 KB it would be without that shared registration. The next structural improvement is to extract serializer registration into a dedicated small VM prelude; omitting it would make replay deserialization incorrect.

The 19 split bundles total 5.28 MB decoded versus the old 1.31 MB monolith because shared sandbox dependencies and the hybrid serializer source repeat. Opaque base64 transport is 7.04 MB on disk before compression (1.33× decoded; gzip is close to the original text). This is a build/storage tradeoff: a replay imports, decodes, compiles, and evaluates only its selected source bundle, and the route caches that decoded promise.

Deployment benchmark

The final benchmark run completed successfully on deployment dpl_CHJqdRSi52QvbKm1KzXhEs4Aa1B2; the sticky comparison reports:

• inline STSO p75 191 → 169 ms (-12%), p90 229 → 191 ms (-17%), p99 580 → 268 ms (-54%)
• whole-run overhead for 1,020 steps 195,405 → 167,186 ms (-14%)
• an earlier run of the same runtime code measured STSO p75/p90/p99 at 155/172/264 ms and whole-run overhead at 153,275 ms, so the improvement survives expected cold-start and infrastructure variance
• TTFS p75 remains 1.12–1.50 s across scenarios, showing that dispatch, route cold start, and workflow-server work still dominate end-to-end startup

A final cold compile-miss trace has a 22.52 ms workflow.run; the surrounding fresh-replay path shows 6.88 ms bundle load (overlapped with 2.37 ms world init), followed by 3.00 ms context creation, 4.00 ms compile, 2.63 ms evaluate, 1.53 ms input hydrate, and 8.82 ms replay execute. Its 146 ms payload-preparation wall span begins before the replay and overlaps event loading and execution rather than representing 146 ms of blocking CPU work.

After the bundle and script caches are warm, the final sequential benchmark's first fresh VM replay trace is 6.88 ms: context creation 2.00 ms, compile hit 0.14 ms, evaluation 1.94 ms, input hydration 0.12 ms, and replay execution 2.41 ms. The comparable main trace is 186.10 ms (-96.3% for this representative warm-cache replay). Subsequent retained invocations in the final trace are generally 1–2.4 ms.

Validation

• 295 focused tests across builders, runtime, tracing, payload cache, and stream recovery
• complete builders suite: 226/226
• affected package typechecks and builds
• Next/Turbopack, Vite, and Nuxt production builds
• exact Next HMR regression plus same-size/same-mtime workflow-ID rename regression
• all eight stable/canary × Webpack/Turbopack × Node/QuickJS Next dev E2E lanes green
• Linux and Windows unit lanes plus Node/QuickJS Windows Next E2E lanes green
• standalone/BOPA import smoke test
• final benchmark workflow green
• final Codex and Claude autoreviews clean

[changeset-bot]

🦋 Changeset detected

Latest commit: 770e721

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 16 packages

Name	Type
@workflow/builders	Patch
@workflow/core	Patch
@workflow/next	Patch
@workflow/world-testing	Patch
workflow	Patch
@workflow/astro	Patch
@workflow/cli	Patch
@workflow/nest	Patch
@workflow/nitro	Patch
@workflow/rollup	Patch
@workflow/sveltekit	Patch
@workflow/vite	Patch
@workflow/vitest	Patch
@workflow/web-shared	Patch
@workflow/web	Patch
@workflow/nuxt	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-express-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nestjs-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-python-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-docs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-swc-playground	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-tarballs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-web	Ready	Preview, v0	Sep 1, 2026 3:32am UTC

[github-actions]

🧪 E2E Test Results

❌ Some tests failed

❌ Failed E2E Tests

▲ Vercel Production (8 failed)

python-node (8 failed):

• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3 | 🔍 observability
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M | 🔍 observability
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23 | 🔍 observability
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG | 🔍 observability
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115 | 🔍 observability
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP | 🔍 observability
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY | 🔍 observability
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF | 🔍 observability

🌐 Cross-language Conformance (9 failed)

python (9 failed):

• deploymentId: 'latest' is a no-op in non-Vercel worlds | wrun_01M1DGN5VWSB6FR072SCMVKD93
• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

• cancelRun via CLI - cancelling a running workflow (nextjs-turbopack)
• cancelRun via CLI - cancelling a running workflow (nuxt)
• promiseAllWorkflow (nest)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

31 infra events

• cold-start-warmup · suite warmup (python) · at 03:32:38Z · abandoned wrun_41M1DG9NKP0GT49DFD2FT24FQP · (+7 more)
• run-pickup-stall · nullByteWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA4A0GWNBVG1J3PNW2BM
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F7
• run-pickup-stall · sleepingWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F6
• run-pickup-stall · promiseAllWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA410GG3J9DQS5J8V5RX
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDACG0GYN2PV31FE1722G
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:33:26Z · abandoned wrun_41M1DGE9JE0GYK5Q2XK9JWVR87
• run-pickup-stall · sleepingWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5DP0GZT0NGX0185G3X8
• run-pickup-stall · promiseAllWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5BE0GMZ14CQP9SC1Y7K
• run-pickup-stall · nullByteWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5AD0GMMXJ2AJ6AZN3XB
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:33:55Z · abandoned wrun_41M1DGF5BM0GJKW6R6A11PP5G5
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:34:45Z · abandoned wrun_41M1DGF8SE0GTQDHMN40HTGC3X
• cold-start-warmup · suite warmup (python) · at 03:35:40Z · abandoned wrun_01M1DGF7EGPY04C8PJ02V841CN · (+7 more)
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJV4MD7KSP260W7ZSBC
• run-pickup-stall · sleepingWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK0G2QC0ZT6CH9PZYC0
• run-pickup-stall · promiseAllWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJW8G1Y4FFC31VK77KY
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK38F43AA3AECZXYP9H
• run-pickup-stall · nullByteWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK5160SZJ5HM2H6FMB9
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:36:26Z · abandoned wrun_41M1DGH6TB0GZCNX8E7RCMJJ6N
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:36:26Z · abandoned wrun_41M1DGJ0P00GPSJ9R5XE1S1WHX
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6HAQR493S719104TSK
• run-pickup-stall · sleepingWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6NBAJ7B203HQDK5N2N
• run-pickup-stall · promiseAllWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6KR8J5R749MV6Q056B
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6SZKHYHZ67N6HHVCTV
• run-pickup-stall · nullByteWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ70B63HA9BG628J3E8C
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHT45WP6FSWB7FGZRVDY
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTEREXVJRZ204H80RN5
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTKKDTFSEZV6GQKNE3R
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF60XCRG4QP1WMRPCM2K
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF5YJJWK6M7RQPW03HKJ
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:38:55Z · abandoned wrun_01M1DGRCDP4DNKXYYTR30VTRPT

E2E Test Summary

Summary

	Passed	Failed	Skipped	Total
❌ ▲ Vercel Production	3570	8	742	4320
✅ 💻 Local Development	3922	0	558	4480
✅ 📦 Local Production	3922	0	558	4480
✅ 🐘 Local Postgres	3922	0	558	4480
✅ 🪟 Windows	320	0	0	320
❌ 🌐 Cross-language Conformance	0	9	132	141
✅ vercel-http-transport	817	0	143	960
✅ vercel-multi-region	27	0	0	27
✅ vercel-ws-transport	553	0	87	640
Total	17053	17	2778	19848

Details by Category

❌ ▲ Vercel Production

App	Passed	Failed	Skipped
✅ astro-node	132	0	28
✅ astro-quickjs	132	0	28
✅ example-node	132	0	28
✅ example-quickjs	132	0	28
✅ express-node	132	0	28
✅ express-quickjs	132	0	28
✅ fastify-node	132	0	28
✅ fastify-quickjs	132	0	28
✅ hono-node	132	0	28
✅ hono-quickjs	132	0	28
✅ nest-node	132	0	28
✅ nest-quickjs	132	0	28
✅ nextjs-turbopack-node	157	0	3
✅ nextjs-turbopack-quickjs	157	0	3
✅ nextjs-webpack-node	157	0	3
✅ nextjs-webpack-quickjs	157	0	3
✅ nitro-node	132	0	28
✅ nitro-quickjs	132	0	28
✅ nuxt-node	132	0	28
✅ nuxt-quickjs	132	0	28
❌ python-node	0	8	152
✅ sveltekit-node	151	0	9
✅ sveltekit-quickjs	151	0	9
✅ tanstack-start-node	132	0	28
✅ tanstack-start-quickjs	132	0	28
✅ vite-node	132	0	28
✅ vite-quickjs	132	0	28

✅ 💻 Local Development

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 📦 Local Production

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🐘 Local Postgres

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	160	0	0
✅ nextjs-turbopack-quickjs	160	0	0

❌ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
❌ python	0	9	132

✅ vercel-http-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ hono	132	0	28
✅ nextjs-turbopack	157	0	3
✅ nitro	132	0	28
✅ vite	132	0	28

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ nextjs-turbopack	157	0	3
✅ vite	132	0	28

---

📋 View full workflow run

[github-actions]

📊 Workflow Benchmarks

commit 770e721 · Tue, 01 Sep 2026 03:46:20 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	1102 (+502%) 🔻	1224 🔴 (+15%) 🔻	1258 🔴 (+13%)	1306 🔴 (-8.5%)	30
TTFS	stream	112 (-48%) 💚	1187 🔴 (+9.2%)	1225 🔴 (+12%)	1243 🔴 (+6.9%)	30
TTFS	hook + stream	377 (-69%) 💚	1463 🔴 (+9.8%)	1836 🔴 (+32%) 🔻	2132 🔴 (+36%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	508 (-15%) 💚	1679 (+5.7%)	1705 (-3.1%)	1752 (-6.7%)	10
Fan-out TTLS	Promise.all(100 steps)	1590 (-4.6%)	3569 (+6.8%)	3692 (+4.0%)	4409 (-34%) 💚	10
STSO	1020 steps (inline)	67 (-40%) 💚	133 (-10%)	155 (-11%)	230 (-35%) 💚	1019
WO	1020 steps	131497 (-12%)	131497 (-12%)	131497 (-12%)	131497 (-12%)	1
CRTT	first chunk (pooled)	88 (-9.3%)	120 (-20%) 💚	178 (+3.5%)	355 (+39%) 🔻	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	110 (-25%)	120 (-33%)	185 (-52%)	315 (-58%)	116 (-40%)	10
size sweep (100/s, 160B-12KB)	106 (-17%)	132 (-29%)	268 (-21%)	767 (+62%)	133 (-13%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	117 (-11%)	110 (-31%)	130 (-48%)	224 (-63%)	251 (-24%)	3
replay eve-gpt-5.6-sol-2000t (1x)	115 (-32%)	121 (-29%)	150 (-37%)	335 (-61%)	392 (-36%)	2
replay eve-gpt-5.6-sol-2000t (2x)	120 (-6%)	149 (-38%)	225 (-36%)	457 (-28%)	254 (-11%)	3

📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 148790ms → this run 131259ms (Δ -17531ms, -12%)

 50-100 ms ┃ main 0 this 3 +3
100-150 ms █████████████████████░░┃ main 779 this 872 +93
150-200 ms ███┃█ main 181 this 128 -53
200-250 ms ┃ main 31 this 9 -22
250-300 ms ┃ main 11 this 6 -5
300-350 ms ┃ main 6 this 0 -6
350-400 ms ┃ main 6 this 0 -6
400-450 ms ┃ main 1 this 1 +0
450-500 ms ┃ main 1 this 0 -1
500-550 ms ┃ main 1 this 0 -1
600-650 ms ┃ main 1 this 0 -1
650-700 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)

variant RTT 1ms→5s+ avg p50 p90 p99 n
control ·····▁█▇▁···· 103.2 (-32%) 96 (-24%) 185 (-52%) 315 (-58%) 3000
sweep ······▆█▁▁··· 126.4 (-17%) 107 (-15%) 268 (-21%) 767 (+62%) 3000
gw 1x ·····▁█▅▁···· 95.8 (-30%) 92 (-23%) 130 (-48%) 224 (-63%) 5295
eve 1x ·····▁█▇▁▁··· 103.6 (-35%) 94 (-27%) 150 (-37%) 335 (-61%) 5186
eve 2x ·····▁▅█▂▁··· 128.1 (-32%) 113 (-31%) 225 (-36%) 457 (-28%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▂▃▂▁▅▂▁▃▃ 95–124ms
sweep █▅▃▂▂▄▂▄▂▁ 105–169ms
gw 1x ▃█▃▄▂▂▂▁▂▄ 91–107ms
eve 1x ▃▂▃▃▁█▃▄▂▁ 86–146ms
eve 2x ▂▃▃▁▁▃▆█▅▂ 103–175ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep ▅█▆▃▂▁▂ 123–131ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▅▃▄█▃▅▅▁▃▆ 26–35ms
sweep ▅▆▇▄▇▄▆█▄▁ 41–56ms
gw 1x ▁█▃▃▃▄▆▅▄▅ 23–32ms
eve 1x ▂▃▁▁▂█▅▅▃▃ 19–29ms
eve 2x ▂▄▁▁▁▂█▄▂▆ 19–26ms

ℹ️ Metric definitions & methodology

_{Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.}

_{The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.}

_{The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.}

_{Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.}

_{Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)}

_{Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles}

_{Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600}

_{All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).}

_{Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.}

[github-actions]

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenario	outcome	events	virt	replay	violations
✅ smoke-no-steps	completed	3	0ms	ok	0
✅ smoke-one-step	completed	6	0ms	ok	0
✅ hook-at-step-started	completed	12	0ms	ok	0
✅ hook-at-step-completed	completed	12	0ms	ok	0
✅ hook-at-hook-created	completed	12	0ms	ok	0
✅ deadline-hook-wins	completed	7	1.0h	ok	0
✅ deadline-expires	completed	7	1.0h	ok	0
✅ long-sleep	completed	11	30.0d	ok	0
✅ hook-never-arrives	stalled	3	0ms	skipped	0
✅ step-retries-twice	completed	10	2.0s	ok	0
✅ parallel-steps	completed	9	0ms	ok	0
✅ hook-on-execution-state	completed	12	0ms	ok	0
✅ peek-hook-before-branch	completed	12	0ms	ok	0
✅ peek-hook-after-branch	completed	12	0ms	ok	0
✅ peek-hook-at-registration	completed	12	0ms	ok	0
✅ race-hook-before-probe	completed	12	0ms	ok	0
✅ race-hook-after-probe	completed	12	0ms	ok	0
✅ race-duplicate-delivery	completed	13	0ms	ok	0
✅ attr-hook-before-step	completed	11	0ms	ok	0
✅ attr-hook-after-step	completed	11	0ms	ok	0
✅ attr-from-step-body	completed	13	0ms	ok	0
✅ fork-hook-after-timeout	completed	14	1.0m	ok	0
✅ fork-hook-before-timeout	completed	14	1.0m	ok	0
✅ count-hook-after-timeout	completed	17	1.0m	ok	0
✅ count-hook-before-timeout	completed	20	1.0m	ok	0
✅ stale-read-step-count-fork	completed	20	1.0m	ok	0
✅ stale-read-equal-step-counts	completed	14	1.0m	ok	0
✅ step-vs-step-fork	completed	12	0ms	ok	0
✅ step-vs-step-fork-fenced	completed	12	0ms	ok	0
✅ fence-catches-benign-direction	completed	12	5ms	ok	0
✅ in-flight-before-decision	completed	17	1.0m	ok	0
❌ in-flight-before-decision-counted	completed	17	1.0m	ok	0
✅ in-flight-after-decision	completed	19	2.0m	ok	0
✅ stale-read-step-count-fork-fenced	completed	20	1.0m	ok	0
✅ fork-hook-wins	completed	13	1.0m	ok	0
✅ fork-timeout-wins	completed	13	1.0m	ok	0
✅ unclaimed-payload-under-fork	completed	17	1.0m	ok	0
✅ claimed-payload-under-fork	completed	17	1.0m	ok	0
✅ writers-independent-step-bodies	completed	12	0ms	ok	0
✅ writers-scripted-tempo	completed	12	0ms	ok	0
✅ cancel-mid-step	cancelled	7	0ms	skipped	0

Full trace: world-sim.txt

[vercel]

Additional Suggestion:

Test asserts SPEC_VERSION_CURRENT === SPEC_VERSION_SUPPORTS_COMPRESSION (5) but the PR bumped SPEC_VERSION_CURRENT to SPEC_VERSION_SUPPORTS_SLOT_IDENTITY (6), so the assertion evaluates 6 === 5 and fails.

[github-actions]

Framework	Cold replay	Step reg.	Framework output
hono	304.7 KiB (+103.8 KiB) ⚠️	40.6 KiB (±0)	2.58 MiB (+830.6 KiB)
nextjs-turbopack	305.0 KiB (+98.7 KiB) ⚠️	439 B (±0)	562.1 KiB (-203.4 KiB)

About these numbers

Sizes are gzip; parentheses show the change against main.
Cold replay and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.
⚠️ marks growth past the threshold. Add the allow-bundle-size-growth label to accept it.

770e721 · run

[NathanColosimo]

High Level

Change build output to be separate workflow bundles and load only the needed code in the runtime.
There's very few changes to the runtime code, most of the changes are in the builder code.
Almost all of the changes are tests

Summary

• emit one VM bundle per workflow source; workflows in the same source share one bundle
• keep one flow route with a static workflow-ID → dynamic-import loader map
• overlap the selected bundle load with world initialization and event setup
• skip bundle loading for queue-only step deliveries
• retain compiled scripts for every immutable production source bundle
• keep watch loaders cache-safe and derive loader IDs from the same transform that emitted each bundle
• encode generated VM modules as opaque artifacts so framework plugins cannot rewrite inert source

Result

For the Next/Turbopack benchmark app's 97_bench source, measured against the monolithic bundle at the base of this PR:

• generated flow route: 1,327,514 B → 18,264 B (-98.6%)
• selected benchmark VM code: 1,307,595 B → 133,859 B (-89.8%)
• cold compile p50: 10.405 ms → 1.089 ms (-89.5%)
• fresh-context evaluation p50: 7.484 ms → 0.356 ms (-95.2%)
• one-time base64 artifact decode p50/p90: 0.037/0.044 ms

The compile/evaluation microbenchmarks use 60 uncached vm.Script compilations and 80 evaluations in fresh VM contexts on Node. The deployment benchmark and APM traces remain the end-to-end signal.

The correctness-preserving split deliberately includes any source that defines both workflows and custom serializers in every other source bundle. The workbench's unusually large 99_e2e.ts is such a hybrid, so the selected benchmark bundle is ~134 KB rather than the ~28 KB it would be without that shared registration. The next structural improvement is to extract serializer registration into a dedicated small VM prelude; omitting it would make replay deserialization incorrect.

The 19 split bundles total 5.28 MB decoded versus the old 1.31 MB monolith because shared sandbox dependencies and the hybrid serializer source repeat. Opaque base64 transport is 7.04 MB on disk before compression (1.33× decoded; gzip is close to the original text). This is a build/storage tradeoff: a replay imports, decodes, compiles, and evaluates only its selected source bundle, and the route caches that decoded promise.

Deployment benchmark

The final benchmark run completed successfully on deployment dpl_CHJqdRSi52QvbKm1KzXhEs4Aa1B2; the sticky comparison reports:

• inline STSO p75 191 → 169 ms (-12%), p90 229 → 191 ms (-17%), p99 580 → 268 ms (-54%)
• whole-run overhead for 1,020 steps 195,405 → 167,186 ms (-14%)
• an earlier run of the same runtime code measured STSO p75/p90/p99 at 155/172/264 ms and whole-run overhead at 153,275 ms, so the improvement survives expected cold-start and infrastructure variance
• TTFS p75 remains 1.12–1.50 s across scenarios, showing that dispatch, route cold start, and workflow-server work still dominate end-to-end startup

A final cold compile-miss trace has a 22.52 ms workflow.run; the surrounding fresh-replay path shows 6.88 ms bundle load (overlapped with 2.37 ms world init), followed by 3.00 ms context creation, 4.00 ms compile, 2.63 ms evaluate, 1.53 ms input hydrate, and 8.82 ms replay execute. Its 146 ms payload-preparation wall span begins before the replay and overlaps event loading and execution rather than representing 146 ms of blocking CPU work.

After the bundle and script caches are warm, the final sequential benchmark's first fresh VM replay trace is 6.88 ms: context creation 2.00 ms, compile hit 0.14 ms, evaluation 1.94 ms, input hydration 0.12 ms, and replay execution 2.41 ms. The comparable main trace is 186.10 ms (-96.3% for this representative warm-cache replay). Subsequent retained invocations in the final trace are generally 1–2.4 ms.

Validation

• 295 focused tests across builders, runtime, tracing, payload cache, and stream recovery
• complete builders suite: 226/226
• affected package typechecks and builds
• Next/Turbopack, Vite, and Nuxt production builds
• exact Next HMR regression plus same-size/same-mtime workflow-ID rename regression
• all eight stable/canary × Webpack/Turbopack × Node/QuickJS Next dev E2E lanes green
• Linux and Windows unit lanes plus Node/QuickJS Windows Next E2E lanes green
• standalone/BOPA import smoke test
• final benchmark workflow green
• final Codex and Claude autoreviews clean

[changeset-bot]

🦋 Changeset detected

Latest commit: 770e721

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 16 packages

Name	Type
@workflow/builders	Patch
@workflow/core	Patch
@workflow/next	Patch
@workflow/world-testing	Patch
workflow	Patch
@workflow/astro	Patch
@workflow/cli	Patch
@workflow/nest	Patch
@workflow/nitro	Patch
@workflow/rollup	Patch
@workflow/sveltekit	Patch
@workflow/vite	Patch
@workflow/vitest	Patch
@workflow/web-shared	Patch
@workflow/web	Patch
@workflow/nuxt	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-express-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nestjs-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-python-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-docs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-swc-playground	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-tarballs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-web	Ready	Preview, v0	Sep 1, 2026 3:32am UTC

[github-actions]

🧪 E2E Test Results

❌ Some tests failed

❌ Failed E2E Tests

▲ Vercel Production (8 failed)

python-node (8 failed):

• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3 | 🔍 observability
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M | 🔍 observability
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23 | 🔍 observability
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG | 🔍 observability
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115 | 🔍 observability
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP | 🔍 observability
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY | 🔍 observability
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF | 🔍 observability

🌐 Cross-language Conformance (9 failed)

python (9 failed):

• deploymentId: 'latest' is a no-op in non-Vercel worlds | wrun_01M1DGN5VWSB6FR072SCMVKD93
• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

• cancelRun via CLI - cancelling a running workflow (nextjs-turbopack)
• cancelRun via CLI - cancelling a running workflow (nuxt)
• promiseAllWorkflow (nest)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

31 infra events

• cold-start-warmup · suite warmup (python) · at 03:32:38Z · abandoned wrun_41M1DG9NKP0GT49DFD2FT24FQP · (+7 more)
• run-pickup-stall · nullByteWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA4A0GWNBVG1J3PNW2BM
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F7
• run-pickup-stall · sleepingWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F6
• run-pickup-stall · promiseAllWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA410GG3J9DQS5J8V5RX
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDACG0GYN2PV31FE1722G
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:33:26Z · abandoned wrun_41M1DGE9JE0GYK5Q2XK9JWVR87
• run-pickup-stall · sleepingWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5DP0GZT0NGX0185G3X8
• run-pickup-stall · promiseAllWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5BE0GMZ14CQP9SC1Y7K
• run-pickup-stall · nullByteWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5AD0GMMXJ2AJ6AZN3XB
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:33:55Z · abandoned wrun_41M1DGF5BM0GJKW6R6A11PP5G5
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:34:45Z · abandoned wrun_41M1DGF8SE0GTQDHMN40HTGC3X
• cold-start-warmup · suite warmup (python) · at 03:35:40Z · abandoned wrun_01M1DGF7EGPY04C8PJ02V841CN · (+7 more)
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJV4MD7KSP260W7ZSBC
• run-pickup-stall · sleepingWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK0G2QC0ZT6CH9PZYC0
• run-pickup-stall · promiseAllWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJW8G1Y4FFC31VK77KY
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK38F43AA3AECZXYP9H
• run-pickup-stall · nullByteWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK5160SZJ5HM2H6FMB9
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:36:26Z · abandoned wrun_41M1DGH6TB0GZCNX8E7RCMJJ6N
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:36:26Z · abandoned wrun_41M1DGJ0P00GPSJ9R5XE1S1WHX
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6HAQR493S719104TSK
• run-pickup-stall · sleepingWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6NBAJ7B203HQDK5N2N
• run-pickup-stall · promiseAllWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6KR8J5R749MV6Q056B
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6SZKHYHZ67N6HHVCTV
• run-pickup-stall · nullByteWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ70B63HA9BG628J3E8C
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHT45WP6FSWB7FGZRVDY
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTEREXVJRZ204H80RN5
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTKKDTFSEZV6GQKNE3R
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF60XCRG4QP1WMRPCM2K
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF5YJJWK6M7RQPW03HKJ
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:38:55Z · abandoned wrun_01M1DGRCDP4DNKXYYTR30VTRPT

E2E Test Summary

Summary

	Passed	Failed	Skipped	Total
❌ ▲ Vercel Production	3570	8	742	4320
✅ 💻 Local Development	3922	0	558	4480
✅ 📦 Local Production	3922	0	558	4480
✅ 🐘 Local Postgres	3922	0	558	4480
✅ 🪟 Windows	320	0	0	320
❌ 🌐 Cross-language Conformance	0	9	132	141
✅ vercel-http-transport	817	0	143	960
✅ vercel-multi-region	27	0	0	27
✅ vercel-ws-transport	553	0	87	640
Total	17053	17	2778	19848

Details by Category

❌ ▲ Vercel Production

App	Passed	Failed	Skipped
✅ astro-node	132	0	28
✅ astro-quickjs	132	0	28
✅ example-node	132	0	28
✅ example-quickjs	132	0	28
✅ express-node	132	0	28
✅ express-quickjs	132	0	28
✅ fastify-node	132	0	28
✅ fastify-quickjs	132	0	28
✅ hono-node	132	0	28
✅ hono-quickjs	132	0	28
✅ nest-node	132	0	28
✅ nest-quickjs	132	0	28
✅ nextjs-turbopack-node	157	0	3
✅ nextjs-turbopack-quickjs	157	0	3
✅ nextjs-webpack-node	157	0	3
✅ nextjs-webpack-quickjs	157	0	3
✅ nitro-node	132	0	28
✅ nitro-quickjs	132	0	28
✅ nuxt-node	132	0	28
✅ nuxt-quickjs	132	0	28
❌ python-node	0	8	152
✅ sveltekit-node	151	0	9
✅ sveltekit-quickjs	151	0	9
✅ tanstack-start-node	132	0	28
✅ tanstack-start-quickjs	132	0	28
✅ vite-node	132	0	28
✅ vite-quickjs	132	0	28

✅ 💻 Local Development

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 📦 Local Production

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🐘 Local Postgres

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	160	0	0
✅ nextjs-turbopack-quickjs	160	0	0

❌ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
❌ python	0	9	132

✅ vercel-http-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ hono	132	0	28
✅ nextjs-turbopack	157	0	3
✅ nitro	132	0	28
✅ vite	132	0	28

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ nextjs-turbopack	157	0	3
✅ vite	132	0	28

---

📋 View full workflow run

[github-actions]

📊 Workflow Benchmarks

commit 770e721 · Tue, 01 Sep 2026 03:46:20 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	1102 (+502%) 🔻	1224 🔴 (+15%) 🔻	1258 🔴 (+13%)	1306 🔴 (-8.5%)	30
TTFS	stream	112 (-48%) 💚	1187 🔴 (+9.2%)	1225 🔴 (+12%)	1243 🔴 (+6.9%)	30
TTFS	hook + stream	377 (-69%) 💚	1463 🔴 (+9.8%)	1836 🔴 (+32%) 🔻	2132 🔴 (+36%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	508 (-15%) 💚	1679 (+5.7%)	1705 (-3.1%)	1752 (-6.7%)	10
Fan-out TTLS	Promise.all(100 steps)	1590 (-4.6%)	3569 (+6.8%)	3692 (+4.0%)	4409 (-34%) 💚	10
STSO	1020 steps (inline)	67 (-40%) 💚	133 (-10%)	155 (-11%)	230 (-35%) 💚	1019
WO	1020 steps	131497 (-12%)	131497 (-12%)	131497 (-12%)	131497 (-12%)	1
CRTT	first chunk (pooled)	88 (-9.3%)	120 (-20%) 💚	178 (+3.5%)	355 (+39%) 🔻	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	110 (-25%)	120 (-33%)	185 (-52%)	315 (-58%)	116 (-40%)	10
size sweep (100/s, 160B-12KB)	106 (-17%)	132 (-29%)	268 (-21%)	767 (+62%)	133 (-13%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	117 (-11%)	110 (-31%)	130 (-48%)	224 (-63%)	251 (-24%)	3
replay eve-gpt-5.6-sol-2000t (1x)	115 (-32%)	121 (-29%)	150 (-37%)	335 (-61%)	392 (-36%)	2
replay eve-gpt-5.6-sol-2000t (2x)	120 (-6%)	149 (-38%)	225 (-36%)	457 (-28%)	254 (-11%)	3

📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 148790ms → this run 131259ms (Δ -17531ms, -12%)

 50-100 ms ┃ main 0 this 3 +3
100-150 ms █████████████████████░░┃ main 779 this 872 +93
150-200 ms ███┃█ main 181 this 128 -53
200-250 ms ┃ main 31 this 9 -22
250-300 ms ┃ main 11 this 6 -5
300-350 ms ┃ main 6 this 0 -6
350-400 ms ┃ main 6 this 0 -6
400-450 ms ┃ main 1 this 1 +0
450-500 ms ┃ main 1 this 0 -1
500-550 ms ┃ main 1 this 0 -1
600-650 ms ┃ main 1 this 0 -1
650-700 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)

variant RTT 1ms→5s+ avg p50 p90 p99 n
control ·····▁█▇▁···· 103.2 (-32%) 96 (-24%) 185 (-52%) 315 (-58%) 3000
sweep ······▆█▁▁··· 126.4 (-17%) 107 (-15%) 268 (-21%) 767 (+62%) 3000
gw 1x ·····▁█▅▁···· 95.8 (-30%) 92 (-23%) 130 (-48%) 224 (-63%) 5295
eve 1x ·····▁█▇▁▁··· 103.6 (-35%) 94 (-27%) 150 (-37%) 335 (-61%) 5186
eve 2x ·····▁▅█▂▁··· 128.1 (-32%) 113 (-31%) 225 (-36%) 457 (-28%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▂▃▂▁▅▂▁▃▃ 95–124ms
sweep █▅▃▂▂▄▂▄▂▁ 105–169ms
gw 1x ▃█▃▄▂▂▂▁▂▄ 91–107ms
eve 1x ▃▂▃▃▁█▃▄▂▁ 86–146ms
eve 2x ▂▃▃▁▁▃▆█▅▂ 103–175ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep ▅█▆▃▂▁▂ 123–131ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▅▃▄█▃▅▅▁▃▆ 26–35ms
sweep ▅▆▇▄▇▄▆█▄▁ 41–56ms
gw 1x ▁█▃▃▃▄▆▅▄▅ 23–32ms
eve 1x ▂▃▁▁▂█▅▅▃▃ 19–29ms
eve 2x ▂▄▁▁▁▂█▄▂▆ 19–26ms

ℹ️ Metric definitions & methodology

_{Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.}

_{The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.}

_{The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.}

_{Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.}

_{Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)}

_{Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles}

_{Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600}

_{All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).}

_{Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.}

[github-actions]

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenario	outcome	events	virt	replay	violations
✅ smoke-no-steps	completed	3	0ms	ok	0
✅ smoke-one-step	completed	6	0ms	ok	0
✅ hook-at-step-started	completed	12	0ms	ok	0
✅ hook-at-step-completed	completed	12	0ms	ok	0
✅ hook-at-hook-created	completed	12	0ms	ok	0
✅ deadline-hook-wins	completed	7	1.0h	ok	0
✅ deadline-expires	completed	7	1.0h	ok	0
✅ long-sleep	completed	11	30.0d	ok	0
✅ hook-never-arrives	stalled	3	0ms	skipped	0
✅ step-retries-twice	completed	10	2.0s	ok	0
✅ parallel-steps	completed	9	0ms	ok	0
✅ hook-on-execution-state	completed	12	0ms	ok	0
✅ peek-hook-before-branch	completed	12	0ms	ok	0
✅ peek-hook-after-branch	completed	12	0ms	ok	0
✅ peek-hook-at-registration	completed	12	0ms	ok	0
✅ race-hook-before-probe	completed	12	0ms	ok	0
✅ race-hook-after-probe	completed	12	0ms	ok	0
✅ race-duplicate-delivery	completed	13	0ms	ok	0
✅ attr-hook-before-step	completed	11	0ms	ok	0
✅ attr-hook-after-step	completed	11	0ms	ok	0
✅ attr-from-step-body	completed	13	0ms	ok	0
✅ fork-hook-after-timeout	completed	14	1.0m	ok	0
✅ fork-hook-before-timeout	completed	14	1.0m	ok	0
✅ count-hook-after-timeout	completed	17	1.0m	ok	0
✅ count-hook-before-timeout	completed	20	1.0m	ok	0
✅ stale-read-step-count-fork	completed	20	1.0m	ok	0
✅ stale-read-equal-step-counts	completed	14	1.0m	ok	0
✅ step-vs-step-fork	completed	12	0ms	ok	0
✅ step-vs-step-fork-fenced	completed	12	0ms	ok	0
✅ fence-catches-benign-direction	completed	12	5ms	ok	0
✅ in-flight-before-decision	completed	17	1.0m	ok	0
❌ in-flight-before-decision-counted	completed	17	1.0m	ok	0
✅ in-flight-after-decision	completed	19	2.0m	ok	0
✅ stale-read-step-count-fork-fenced	completed	20	1.0m	ok	0
✅ fork-hook-wins	completed	13	1.0m	ok	0
✅ fork-timeout-wins	completed	13	1.0m	ok	0
✅ unclaimed-payload-under-fork	completed	17	1.0m	ok	0
✅ claimed-payload-under-fork	completed	17	1.0m	ok	0
✅ writers-independent-step-bodies	completed	12	0ms	ok	0
✅ writers-scripted-tempo	completed	12	0ms	ok	0
✅ cancel-mid-step	cancelled	7	0ms	skipped	0

Full trace: world-sim.txt

[vercel]

Additional Suggestion:

Test asserts SPEC_VERSION_CURRENT === SPEC_VERSION_SUPPORTS_COMPRESSION (5) but the PR bumped SPEC_VERSION_CURRENT to SPEC_VERSION_SUPPORTS_SLOT_IDENTITY (6), so the assertion evaluates 6 === 5 and fails.

[github-actions]

Framework	Cold replay	Step reg.	Framework output
hono	304.7 KiB (+103.8 KiB) ⚠️	40.6 KiB (±0)	2.58 MiB (+830.6 KiB)
nextjs-turbopack	305.0 KiB (+98.7 KiB) ⚠️	439 B (±0)	562.1 KiB (-203.4 KiB)

About these numbers

Sizes are gzip; parentheses show the change against main.
Cold replay and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.
⚠️ marks growth past the threshold. Add the allow-bundle-size-growth label to accept it.

770e721 · run

[NathanColosimo]

High Level

Change build output to be separate workflow bundles and load only the needed code in the runtime.
There's very few changes to the runtime code, most of the changes are in the builder code.
Almost all of the changes are tests

Summary

• emit one VM bundle per workflow source; workflows in the same source share one bundle
• keep one flow route with a static workflow-ID → dynamic-import loader map
• overlap the selected bundle load with world initialization and event setup
• skip bundle loading for queue-only step deliveries
• retain compiled scripts for every immutable production source bundle
• keep watch loaders cache-safe and derive loader IDs from the same transform that emitted each bundle
• encode generated VM modules as opaque artifacts so framework plugins cannot rewrite inert source

Result

For the Next/Turbopack benchmark app's 97_bench source, measured against the monolithic bundle at the base of this PR:

• generated flow route: 1,327,514 B → 18,264 B (-98.6%)
• selected benchmark VM code: 1,307,595 B → 133,859 B (-89.8%)
• cold compile p50: 10.405 ms → 1.089 ms (-89.5%)
• fresh-context evaluation p50: 7.484 ms → 0.356 ms (-95.2%)
• one-time base64 artifact decode p50/p90: 0.037/0.044 ms

The compile/evaluation microbenchmarks use 60 uncached vm.Script compilations and 80 evaluations in fresh VM contexts on Node. The deployment benchmark and APM traces remain the end-to-end signal.

The correctness-preserving split deliberately includes any source that defines both workflows and custom serializers in every other source bundle. The workbench's unusually large 99_e2e.ts is such a hybrid, so the selected benchmark bundle is ~134 KB rather than the ~28 KB it would be without that shared registration. The next structural improvement is to extract serializer registration into a dedicated small VM prelude; omitting it would make replay deserialization incorrect.

The 19 split bundles total 5.28 MB decoded versus the old 1.31 MB monolith because shared sandbox dependencies and the hybrid serializer source repeat. Opaque base64 transport is 7.04 MB on disk before compression (1.33× decoded; gzip is close to the original text). This is a build/storage tradeoff: a replay imports, decodes, compiles, and evaluates only its selected source bundle, and the route caches that decoded promise.

Deployment benchmark

The final benchmark run completed successfully on deployment dpl_CHJqdRSi52QvbKm1KzXhEs4Aa1B2; the sticky comparison reports:

• inline STSO p75 191 → 169 ms (-12%), p90 229 → 191 ms (-17%), p99 580 → 268 ms (-54%)
• whole-run overhead for 1,020 steps 195,405 → 167,186 ms (-14%)
• an earlier run of the same runtime code measured STSO p75/p90/p99 at 155/172/264 ms and whole-run overhead at 153,275 ms, so the improvement survives expected cold-start and infrastructure variance
• TTFS p75 remains 1.12–1.50 s across scenarios, showing that dispatch, route cold start, and workflow-server work still dominate end-to-end startup

A final cold compile-miss trace has a 22.52 ms workflow.run; the surrounding fresh-replay path shows 6.88 ms bundle load (overlapped with 2.37 ms world init), followed by 3.00 ms context creation, 4.00 ms compile, 2.63 ms evaluate, 1.53 ms input hydrate, and 8.82 ms replay execute. Its 146 ms payload-preparation wall span begins before the replay and overlaps event loading and execution rather than representing 146 ms of blocking CPU work.

After the bundle and script caches are warm, the final sequential benchmark's first fresh VM replay trace is 6.88 ms: context creation 2.00 ms, compile hit 0.14 ms, evaluation 1.94 ms, input hydration 0.12 ms, and replay execution 2.41 ms. The comparable main trace is 186.10 ms (-96.3% for this representative warm-cache replay). Subsequent retained invocations in the final trace are generally 1–2.4 ms.

Validation

• 295 focused tests across builders, runtime, tracing, payload cache, and stream recovery
• complete builders suite: 226/226
• affected package typechecks and builds
• Next/Turbopack, Vite, and Nuxt production builds
• exact Next HMR regression plus same-size/same-mtime workflow-ID rename regression
• all eight stable/canary × Webpack/Turbopack × Node/QuickJS Next dev E2E lanes green
• Linux and Windows unit lanes plus Node/QuickJS Windows Next E2E lanes green
• standalone/BOPA import smoke test
• final benchmark workflow green
• final Codex and Claude autoreviews clean

[changeset-bot]

🦋 Changeset detected

Latest commit: 770e721

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 16 packages

Name	Type
@workflow/builders	Patch
@workflow/core	Patch
@workflow/next	Patch
@workflow/world-testing	Patch
workflow	Patch
@workflow/astro	Patch
@workflow/cli	Patch
@workflow/nest	Patch
@workflow/nitro	Patch
@workflow/rollup	Patch
@workflow/sveltekit	Patch
@workflow/vite	Patch
@workflow/vitest	Patch
@workflow/web-shared	Patch
@workflow/web	Patch
@workflow/nuxt	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-express-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nestjs-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-python-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-docs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-swc-playground	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-tarballs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-web	Ready	Preview, v0	Sep 1, 2026 3:32am UTC

[github-actions]

🧪 E2E Test Results

❌ Some tests failed

❌ Failed E2E Tests

▲ Vercel Production (8 failed)

python-node (8 failed):

• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3 | 🔍 observability
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M | 🔍 observability
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23 | 🔍 observability
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG | 🔍 observability
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115 | 🔍 observability
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP | 🔍 observability
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY | 🔍 observability
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF | 🔍 observability

🌐 Cross-language Conformance (9 failed)

python (9 failed):

• deploymentId: 'latest' is a no-op in non-Vercel worlds | wrun_01M1DGN5VWSB6FR072SCMVKD93
• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

• cancelRun via CLI - cancelling a running workflow (nextjs-turbopack)
• cancelRun via CLI - cancelling a running workflow (nuxt)
• promiseAllWorkflow (nest)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

31 infra events

• cold-start-warmup · suite warmup (python) · at 03:32:38Z · abandoned wrun_41M1DG9NKP0GT49DFD2FT24FQP · (+7 more)
• run-pickup-stall · nullByteWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA4A0GWNBVG1J3PNW2BM
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F7
• run-pickup-stall · sleepingWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F6
• run-pickup-stall · promiseAllWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA410GG3J9DQS5J8V5RX
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDACG0GYN2PV31FE1722G
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:33:26Z · abandoned wrun_41M1DGE9JE0GYK5Q2XK9JWVR87
• run-pickup-stall · sleepingWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5DP0GZT0NGX0185G3X8
• run-pickup-stall · promiseAllWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5BE0GMZ14CQP9SC1Y7K
• run-pickup-stall · nullByteWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5AD0GMMXJ2AJ6AZN3XB
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:33:55Z · abandoned wrun_41M1DGF5BM0GJKW6R6A11PP5G5
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:34:45Z · abandoned wrun_41M1DGF8SE0GTQDHMN40HTGC3X
• cold-start-warmup · suite warmup (python) · at 03:35:40Z · abandoned wrun_01M1DGF7EGPY04C8PJ02V841CN · (+7 more)
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJV4MD7KSP260W7ZSBC
• run-pickup-stall · sleepingWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK0G2QC0ZT6CH9PZYC0
• run-pickup-stall · promiseAllWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJW8G1Y4FFC31VK77KY
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK38F43AA3AECZXYP9H
• run-pickup-stall · nullByteWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK5160SZJ5HM2H6FMB9
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:36:26Z · abandoned wrun_41M1DGH6TB0GZCNX8E7RCMJJ6N
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:36:26Z · abandoned wrun_41M1DGJ0P00GPSJ9R5XE1S1WHX
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6HAQR493S719104TSK
• run-pickup-stall · sleepingWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6NBAJ7B203HQDK5N2N
• run-pickup-stall · promiseAllWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6KR8J5R749MV6Q056B
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6SZKHYHZ67N6HHVCTV
• run-pickup-stall · nullByteWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ70B63HA9BG628J3E8C
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHT45WP6FSWB7FGZRVDY
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTEREXVJRZ204H80RN5
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTKKDTFSEZV6GQKNE3R
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF60XCRG4QP1WMRPCM2K
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF5YJJWK6M7RQPW03HKJ
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:38:55Z · abandoned wrun_01M1DGRCDP4DNKXYYTR30VTRPT

E2E Test Summary

Summary

	Passed	Failed	Skipped	Total
❌ ▲ Vercel Production	3570	8	742	4320
✅ 💻 Local Development	3922	0	558	4480
✅ 📦 Local Production	3922	0	558	4480
✅ 🐘 Local Postgres	3922	0	558	4480
✅ 🪟 Windows	320	0	0	320
❌ 🌐 Cross-language Conformance	0	9	132	141
✅ vercel-http-transport	817	0	143	960
✅ vercel-multi-region	27	0	0	27
✅ vercel-ws-transport	553	0	87	640
Total	17053	17	2778	19848

Details by Category

❌ ▲ Vercel Production

App	Passed	Failed	Skipped
✅ astro-node	132	0	28
✅ astro-quickjs	132	0	28
✅ example-node	132	0	28
✅ example-quickjs	132	0	28
✅ express-node	132	0	28
✅ express-quickjs	132	0	28
✅ fastify-node	132	0	28
✅ fastify-quickjs	132	0	28
✅ hono-node	132	0	28
✅ hono-quickjs	132	0	28
✅ nest-node	132	0	28
✅ nest-quickjs	132	0	28
✅ nextjs-turbopack-node	157	0	3
✅ nextjs-turbopack-quickjs	157	0	3
✅ nextjs-webpack-node	157	0	3
✅ nextjs-webpack-quickjs	157	0	3
✅ nitro-node	132	0	28
✅ nitro-quickjs	132	0	28
✅ nuxt-node	132	0	28
✅ nuxt-quickjs	132	0	28
❌ python-node	0	8	152
✅ sveltekit-node	151	0	9
✅ sveltekit-quickjs	151	0	9
✅ tanstack-start-node	132	0	28
✅ tanstack-start-quickjs	132	0	28
✅ vite-node	132	0	28
✅ vite-quickjs	132	0	28

✅ 💻 Local Development

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 📦 Local Production

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🐘 Local Postgres

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	160	0	0
✅ nextjs-turbopack-quickjs	160	0	0

❌ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
❌ python	0	9	132

✅ vercel-http-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ hono	132	0	28
✅ nextjs-turbopack	157	0	3
✅ nitro	132	0	28
✅ vite	132	0	28

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ nextjs-turbopack	157	0	3
✅ vite	132	0	28

---

📋 View full workflow run

[github-actions]

📊 Workflow Benchmarks

commit 770e721 · Tue, 01 Sep 2026 03:46:20 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	1102 (+502%) 🔻	1224 🔴 (+15%) 🔻	1258 🔴 (+13%)	1306 🔴 (-8.5%)	30
TTFS	stream	112 (-48%) 💚	1187 🔴 (+9.2%)	1225 🔴 (+12%)	1243 🔴 (+6.9%)	30
TTFS	hook + stream	377 (-69%) 💚	1463 🔴 (+9.8%)	1836 🔴 (+32%) 🔻	2132 🔴 (+36%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	508 (-15%) 💚	1679 (+5.7%)	1705 (-3.1%)	1752 (-6.7%)	10
Fan-out TTLS	Promise.all(100 steps)	1590 (-4.6%)	3569 (+6.8%)	3692 (+4.0%)	4409 (-34%) 💚	10
STSO	1020 steps (inline)	67 (-40%) 💚	133 (-10%)	155 (-11%)	230 (-35%) 💚	1019
WO	1020 steps	131497 (-12%)	131497 (-12%)	131497 (-12%)	131497 (-12%)	1
CRTT	first chunk (pooled)	88 (-9.3%)	120 (-20%) 💚	178 (+3.5%)	355 (+39%) 🔻	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	110 (-25%)	120 (-33%)	185 (-52%)	315 (-58%)	116 (-40%)	10
size sweep (100/s, 160B-12KB)	106 (-17%)	132 (-29%)	268 (-21%)	767 (+62%)	133 (-13%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	117 (-11%)	110 (-31%)	130 (-48%)	224 (-63%)	251 (-24%)	3
replay eve-gpt-5.6-sol-2000t (1x)	115 (-32%)	121 (-29%)	150 (-37%)	335 (-61%)	392 (-36%)	2
replay eve-gpt-5.6-sol-2000t (2x)	120 (-6%)	149 (-38%)	225 (-36%)	457 (-28%)	254 (-11%)	3

📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 148790ms → this run 131259ms (Δ -17531ms, -12%)

 50-100 ms ┃ main 0 this 3 +3
100-150 ms █████████████████████░░┃ main 779 this 872 +93
150-200 ms ███┃█ main 181 this 128 -53
200-250 ms ┃ main 31 this 9 -22
250-300 ms ┃ main 11 this 6 -5
300-350 ms ┃ main 6 this 0 -6
350-400 ms ┃ main 6 this 0 -6
400-450 ms ┃ main 1 this 1 +0
450-500 ms ┃ main 1 this 0 -1
500-550 ms ┃ main 1 this 0 -1
600-650 ms ┃ main 1 this 0 -1
650-700 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)

variant RTT 1ms→5s+ avg p50 p90 p99 n
control ·····▁█▇▁···· 103.2 (-32%) 96 (-24%) 185 (-52%) 315 (-58%) 3000
sweep ······▆█▁▁··· 126.4 (-17%) 107 (-15%) 268 (-21%) 767 (+62%) 3000
gw 1x ·····▁█▅▁···· 95.8 (-30%) 92 (-23%) 130 (-48%) 224 (-63%) 5295
eve 1x ·····▁█▇▁▁··· 103.6 (-35%) 94 (-27%) 150 (-37%) 335 (-61%) 5186
eve 2x ·····▁▅█▂▁··· 128.1 (-32%) 113 (-31%) 225 (-36%) 457 (-28%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▂▃▂▁▅▂▁▃▃ 95–124ms
sweep █▅▃▂▂▄▂▄▂▁ 105–169ms
gw 1x ▃█▃▄▂▂▂▁▂▄ 91–107ms
eve 1x ▃▂▃▃▁█▃▄▂▁ 86–146ms
eve 2x ▂▃▃▁▁▃▆█▅▂ 103–175ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep ▅█▆▃▂▁▂ 123–131ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▅▃▄█▃▅▅▁▃▆ 26–35ms
sweep ▅▆▇▄▇▄▆█▄▁ 41–56ms
gw 1x ▁█▃▃▃▄▆▅▄▅ 23–32ms
eve 1x ▂▃▁▁▂█▅▅▃▃ 19–29ms
eve 2x ▂▄▁▁▁▂█▄▂▆ 19–26ms

ℹ️ Metric definitions & methodology

_{Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.}

_{The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.}

_{The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.}

_{Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.}

_{Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)}

_{Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles}

_{Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600}

_{All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).}

_{Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.}

[github-actions]

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenario	outcome	events	virt	replay	violations
✅ smoke-no-steps	completed	3	0ms	ok	0
✅ smoke-one-step	completed	6	0ms	ok	0
✅ hook-at-step-started	completed	12	0ms	ok	0
✅ hook-at-step-completed	completed	12	0ms	ok	0
✅ hook-at-hook-created	completed	12	0ms	ok	0
✅ deadline-hook-wins	completed	7	1.0h	ok	0
✅ deadline-expires	completed	7	1.0h	ok	0
✅ long-sleep	completed	11	30.0d	ok	0
✅ hook-never-arrives	stalled	3	0ms	skipped	0
✅ step-retries-twice	completed	10	2.0s	ok	0
✅ parallel-steps	completed	9	0ms	ok	0
✅ hook-on-execution-state	completed	12	0ms	ok	0
✅ peek-hook-before-branch	completed	12	0ms	ok	0
✅ peek-hook-after-branch	completed	12	0ms	ok	0
✅ peek-hook-at-registration	completed	12	0ms	ok	0
✅ race-hook-before-probe	completed	12	0ms	ok	0
✅ race-hook-after-probe	completed	12	0ms	ok	0
✅ race-duplicate-delivery	completed	13	0ms	ok	0
✅ attr-hook-before-step	completed	11	0ms	ok	0
✅ attr-hook-after-step	completed	11	0ms	ok	0
✅ attr-from-step-body	completed	13	0ms	ok	0
✅ fork-hook-after-timeout	completed	14	1.0m	ok	0
✅ fork-hook-before-timeout	completed	14	1.0m	ok	0
✅ count-hook-after-timeout	completed	17	1.0m	ok	0
✅ count-hook-before-timeout	completed	20	1.0m	ok	0
✅ stale-read-step-count-fork	completed	20	1.0m	ok	0
✅ stale-read-equal-step-counts	completed	14	1.0m	ok	0
✅ step-vs-step-fork	completed	12	0ms	ok	0
✅ step-vs-step-fork-fenced	completed	12	0ms	ok	0
✅ fence-catches-benign-direction	completed	12	5ms	ok	0
✅ in-flight-before-decision	completed	17	1.0m	ok	0
❌ in-flight-before-decision-counted	completed	17	1.0m	ok	0
✅ in-flight-after-decision	completed	19	2.0m	ok	0
✅ stale-read-step-count-fork-fenced	completed	20	1.0m	ok	0
✅ fork-hook-wins	completed	13	1.0m	ok	0
✅ fork-timeout-wins	completed	13	1.0m	ok	0
✅ unclaimed-payload-under-fork	completed	17	1.0m	ok	0
✅ claimed-payload-under-fork	completed	17	1.0m	ok	0
✅ writers-independent-step-bodies	completed	12	0ms	ok	0
✅ writers-scripted-tempo	completed	12	0ms	ok	0
✅ cancel-mid-step	cancelled	7	0ms	skipped	0

Full trace: world-sim.txt

[vercel]

Additional Suggestion:

Test asserts SPEC_VERSION_CURRENT === SPEC_VERSION_SUPPORTS_COMPRESSION (5) but the PR bumped SPEC_VERSION_CURRENT to SPEC_VERSION_SUPPORTS_SLOT_IDENTITY (6), so the assertion evaluates 6 === 5 and fails.

[github-actions]

Framework	Cold replay	Step reg.	Framework output
hono	304.7 KiB (+103.8 KiB) ⚠️	40.6 KiB (±0)	2.58 MiB (+830.6 KiB)
nextjs-turbopack	305.0 KiB (+98.7 KiB) ⚠️	439 B (±0)	562.1 KiB (-203.4 KiB)

About these numbers

Sizes are gzip; parentheses show the change against main.
Cold replay and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.
⚠️ marks growth past the threshold. Add the allow-bundle-size-growth label to accept it.

770e721 · run

[NathanColosimo]

High Level

Change build output to be separate workflow bundles and load only the needed code in the runtime.
There's very few changes to the runtime code, most of the changes are in the builder code.
Almost all of the changes are tests

Summary

• emit one VM bundle per workflow source; workflows in the same source share one bundle
• keep one flow route with a static workflow-ID → dynamic-import loader map
• overlap the selected bundle load with world initialization and event setup
• skip bundle loading for queue-only step deliveries
• retain compiled scripts for every immutable production source bundle
• keep watch loaders cache-safe and derive loader IDs from the same transform that emitted each bundle
• encode generated VM modules as opaque artifacts so framework plugins cannot rewrite inert source

Result

For the Next/Turbopack benchmark app's 97_bench source, measured against the monolithic bundle at the base of this PR:

• generated flow route: 1,327,514 B → 18,264 B (-98.6%)
• selected benchmark VM code: 1,307,595 B → 133,859 B (-89.8%)
• cold compile p50: 10.405 ms → 1.089 ms (-89.5%)
• fresh-context evaluation p50: 7.484 ms → 0.356 ms (-95.2%)
• one-time base64 artifact decode p50/p90: 0.037/0.044 ms

The compile/evaluation microbenchmarks use 60 uncached vm.Script compilations and 80 evaluations in fresh VM contexts on Node. The deployment benchmark and APM traces remain the end-to-end signal.

The correctness-preserving split deliberately includes any source that defines both workflows and custom serializers in every other source bundle. The workbench's unusually large 99_e2e.ts is such a hybrid, so the selected benchmark bundle is ~134 KB rather than the ~28 KB it would be without that shared registration. The next structural improvement is to extract serializer registration into a dedicated small VM prelude; omitting it would make replay deserialization incorrect.

The 19 split bundles total 5.28 MB decoded versus the old 1.31 MB monolith because shared sandbox dependencies and the hybrid serializer source repeat. Opaque base64 transport is 7.04 MB on disk before compression (1.33× decoded; gzip is close to the original text). This is a build/storage tradeoff: a replay imports, decodes, compiles, and evaluates only its selected source bundle, and the route caches that decoded promise.

Deployment benchmark

The final benchmark run completed successfully on deployment dpl_CHJqdRSi52QvbKm1KzXhEs4Aa1B2; the sticky comparison reports:

• inline STSO p75 191 → 169 ms (-12%), p90 229 → 191 ms (-17%), p99 580 → 268 ms (-54%)
• whole-run overhead for 1,020 steps 195,405 → 167,186 ms (-14%)
• an earlier run of the same runtime code measured STSO p75/p90/p99 at 155/172/264 ms and whole-run overhead at 153,275 ms, so the improvement survives expected cold-start and infrastructure variance
• TTFS p75 remains 1.12–1.50 s across scenarios, showing that dispatch, route cold start, and workflow-server work still dominate end-to-end startup

A final cold compile-miss trace has a 22.52 ms workflow.run; the surrounding fresh-replay path shows 6.88 ms bundle load (overlapped with 2.37 ms world init), followed by 3.00 ms context creation, 4.00 ms compile, 2.63 ms evaluate, 1.53 ms input hydrate, and 8.82 ms replay execute. Its 146 ms payload-preparation wall span begins before the replay and overlaps event loading and execution rather than representing 146 ms of blocking CPU work.

After the bundle and script caches are warm, the final sequential benchmark's first fresh VM replay trace is 6.88 ms: context creation 2.00 ms, compile hit 0.14 ms, evaluation 1.94 ms, input hydration 0.12 ms, and replay execution 2.41 ms. The comparable main trace is 186.10 ms (-96.3% for this representative warm-cache replay). Subsequent retained invocations in the final trace are generally 1–2.4 ms.

Validation

• 295 focused tests across builders, runtime, tracing, payload cache, and stream recovery
• complete builders suite: 226/226
• affected package typechecks and builds
• Next/Turbopack, Vite, and Nuxt production builds
• exact Next HMR regression plus same-size/same-mtime workflow-ID rename regression
• all eight stable/canary × Webpack/Turbopack × Node/QuickJS Next dev E2E lanes green
• Linux and Windows unit lanes plus Node/QuickJS Windows Next E2E lanes green
• standalone/BOPA import smoke test
• final benchmark workflow green
• final Codex and Claude autoreviews clean

[changeset-bot]

🦋 Changeset detected

Latest commit: 770e721

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 16 packages

Name	Type
@workflow/builders	Patch
@workflow/core	Patch
@workflow/next	Patch
@workflow/world-testing	Patch
workflow	Patch
@workflow/astro	Patch
@workflow/cli	Patch
@workflow/nest	Patch
@workflow/nitro	Patch
@workflow/rollup	Patch
@workflow/sveltekit	Patch
@workflow/vite	Patch
@workflow/vitest	Patch
@workflow/web-shared	Patch
@workflow/web	Patch
@workflow/nuxt	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
example-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-express-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nestjs-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-python-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-docs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-swc-playground	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-tarballs	Ready	Preview, v0	Sep 1, 2026 3:32am UTC
workflow-web	Ready	Preview, v0	Sep 1, 2026 3:32am UTC

[github-actions]

🧪 E2E Test Results

❌ Some tests failed

❌ Failed E2E Tests

▲ Vercel Production (8 failed)

python-node (8 failed):

• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3 | 🔍 observability
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M | 🔍 observability
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23 | 🔍 observability
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG | 🔍 observability
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115 | 🔍 observability
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP | 🔍 observability
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY | 🔍 observability
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF | 🔍 observability

🌐 Cross-language Conformance (9 failed)

python (9 failed):

• deploymentId: 'latest' is a no-op in non-Vercel worlds | wrun_01M1DGN5VWSB6FR072SCMVKD93
• promiseAllWorkflow | wrun_41M1DGAP800GJNMEJ7SKXS2WR3
• sleepingWorkflow | wrun_41M1DGBBNM0GRR3VT84RPGT37M
• parallelSleepWorkflow | wrun_41M1DGBBX30GJQK8GDEKASRB23
• nullByteWorkflow | wrun_41M1DGBJ7S0GGY055R4SKW2KGG
• cancelRun - cancelling a running workflow | wrun_41M1DGG7830GNN6EDY1MQCQ115
• cancelRun via CLI - cancelling a running workflow | wrun_41M1DGG8BD0GJG1C1TR3XYBKQP
• sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration | wrun_41M1DGGFE30GK2S2QVN4AV27AY
• resilient start: addTenWorkflow completes when run_created returns 500 | wrun_41M1DGH1J20GQM56QKA01JAGWF

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

• cancelRun via CLI - cancelling a running workflow (nextjs-turbopack)
• cancelRun via CLI - cancelling a running workflow (nuxt)
• promiseAllWorkflow (nest)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

31 infra events

• cold-start-warmup · suite warmup (python) · at 03:32:38Z · abandoned wrun_41M1DG9NKP0GT49DFD2FT24FQP · (+7 more)
• run-pickup-stall · nullByteWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA4A0GWNBVG1J3PNW2BM
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F7
• run-pickup-stall · sleepingWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA470GQ7T8KYJDV9H8F6
• run-pickup-stall · promiseAllWorkflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDA410GG3J9DQS5J8V5RX
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:32:54Z · abandoned wrun_41M1DGDACG0GYN2PV31FE1722G
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:33:26Z · abandoned wrun_41M1DGE9JE0GYK5Q2XK9JWVR87
• run-pickup-stall · sleepingWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5DP0GZT0NGX0185G3X8
• run-pickup-stall · promiseAllWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5BE0GMZ14CQP9SC1Y7K
• run-pickup-stall · nullByteWorkflow (python) · at 03:33:54Z · abandoned wrun_41M1DGF5AD0GMMXJ2AJ6AZN3XB
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:33:55Z · abandoned wrun_41M1DGF5BM0GJKW6R6A11PP5G5
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:34:45Z · abandoned wrun_41M1DGF8SE0GTQDHMN40HTGC3X
• cold-start-warmup · suite warmup (python) · at 03:35:40Z · abandoned wrun_01M1DGF7EGPY04C8PJ02V841CN · (+7 more)
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJV4MD7KSP260W7ZSBC
• run-pickup-stall · sleepingWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK0G2QC0ZT6CH9PZYC0
• run-pickup-stall · promiseAllWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWJW8G1Y4FFC31VK77KY
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK38F43AA3AECZXYP9H
• run-pickup-stall · nullByteWorkflow (python) · at 03:35:55Z · abandoned wrun_01M1DGJWK5160SZJ5HM2H6FMB9
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:36:26Z · abandoned wrun_41M1DGH6TB0GZCNX8E7RCMJJ6N
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:36:26Z · abandoned wrun_41M1DGJ0P00GPSJ9R5XE1S1WHX
• run-pickup-stall · deploymentId: 'latest' is a no-op in non-Vercel worlds (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6HAQR493S719104TSK
• run-pickup-stall · sleepingWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6NBAJ7B203HQDK5N2N
• run-pickup-stall · promiseAllWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6KR8J5R749MV6Q056B
• run-pickup-stall · parallelSleepWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ6SZKHYHZ67N6HHVCTV
• run-pickup-stall · nullByteWorkflow (python) · at 03:36:55Z · abandoned wrun_01M1DGMQ70B63HA9BG628J3E8C
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHT45WP6FSWB7FGZRVDY
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTEREXVJRZ204H80RN5
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:37:55Z · abandoned wrun_01M1DGPHTKKDTFSEZV6GQKNE3R
• run-pickup-stall · cancelRun via CLI - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF60XCRG4QP1WMRPCM2K
• run-pickup-stall · cancelRun - cancelling a running workflow (python) · at 03:38:26Z · abandoned wrun_01M1DGQF5YJJWK6M7RQPW03HKJ
• run-pickup-stall · sleepInLoopWorkflow - sleep inside loop with steps actually delays each iteration (python) · at 03:38:55Z · abandoned wrun_01M1DGRCDP4DNKXYYTR30VTRPT

E2E Test Summary

Summary

	Passed	Failed	Skipped	Total
❌ ▲ Vercel Production	3570	8	742	4320
✅ 💻 Local Development	3922	0	558	4480
✅ 📦 Local Production	3922	0	558	4480
✅ 🐘 Local Postgres	3922	0	558	4480
✅ 🪟 Windows	320	0	0	320
❌ 🌐 Cross-language Conformance	0	9	132	141
✅ vercel-http-transport	817	0	143	960
✅ vercel-multi-region	27	0	0	27
✅ vercel-ws-transport	553	0	87	640
Total	17053	17	2778	19848

Details by Category

❌ ▲ Vercel Production

App	Passed	Failed	Skipped
✅ astro-node	132	0	28
✅ astro-quickjs	132	0	28
✅ example-node	132	0	28
✅ example-quickjs	132	0	28
✅ express-node	132	0	28
✅ express-quickjs	132	0	28
✅ fastify-node	132	0	28
✅ fastify-quickjs	132	0	28
✅ hono-node	132	0	28
✅ hono-quickjs	132	0	28
✅ nest-node	132	0	28
✅ nest-quickjs	132	0	28
✅ nextjs-turbopack-node	157	0	3
✅ nextjs-turbopack-quickjs	157	0	3
✅ nextjs-webpack-node	157	0	3
✅ nextjs-webpack-quickjs	157	0	3
✅ nitro-node	132	0	28
✅ nitro-quickjs	132	0	28
✅ nuxt-node	132	0	28
✅ nuxt-quickjs	132	0	28
❌ python-node	0	8	152
✅ sveltekit-node	151	0	9
✅ sveltekit-quickjs	151	0	9
✅ tanstack-start-node	132	0	28
✅ tanstack-start-quickjs	132	0	28
✅ vite-node	132	0	28
✅ vite-quickjs	132	0	28

✅ 💻 Local Development

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 📦 Local Production

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🐘 Local Postgres

App	Passed	Failed	Skipped
✅ astro-stable-node	134	0	26
✅ astro-stable-quickjs	134	0	26
✅ express-stable-node	134	0	26
✅ express-stable-quickjs	134	0	26
✅ fastify-stable-node	134	0	26
✅ fastify-stable-quickjs	134	0	26
✅ hono-stable-node	134	0	26
✅ hono-stable-quickjs	134	0	26
✅ nest-stable-node	134	0	26
✅ nest-stable-quickjs	134	0	26
✅ nextjs-turbopack-canary-node	141	0	19
✅ nextjs-turbopack-canary-quickjs	141	0	19
✅ nextjs-turbopack-stable-node	160	0	0
✅ nextjs-turbopack-stable-quickjs	160	0	0
✅ nextjs-webpack-canary-node	141	0	19
✅ nextjs-webpack-canary-quickjs	141	0	19
✅ nextjs-webpack-stable-node	160	0	0
✅ nextjs-webpack-stable-quickjs	160	0	0
✅ nitro-stable-node	134	0	26
✅ nitro-stable-quickjs	134	0	26
✅ nuxt-stable-node	134	0	26
✅ nuxt-stable-quickjs	134	0	26
✅ sveltekit-stable-node	153	0	7
✅ sveltekit-stable-quickjs	153	0	7
✅ tanstack-start-node	134	0	26
✅ tanstack-start-quickjs	134	0	26
✅ vite-stable-node	134	0	26
✅ vite-stable-quickjs	134	0	26

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	160	0	0
✅ nextjs-turbopack-quickjs	160	0	0

❌ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
❌ python	0	9	132

✅ vercel-http-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ hono	132	0	28
✅ nextjs-turbopack	157	0	3
✅ nitro	132	0	28
✅ vite	132	0	28

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Failed	Skipped
✅ example	132	0	28
✅ express	132	0	28
✅ nextjs-turbopack	157	0	3
✅ vite	132	0	28

---

📋 View full workflow run

[github-actions]

📊 Workflow Benchmarks

commit 770e721 · Tue, 01 Sep 2026 03:46:20 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	1102 (+502%) 🔻	1224 🔴 (+15%) 🔻	1258 🔴 (+13%)	1306 🔴 (-8.5%)	30
TTFS	stream	112 (-48%) 💚	1187 🔴 (+9.2%)	1225 🔴 (+12%)	1243 🔴 (+6.9%)	30
TTFS	hook + stream	377 (-69%) 💚	1463 🔴 (+9.8%)	1836 🔴 (+32%) 🔻	2132 🔴 (+36%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	508 (-15%) 💚	1679 (+5.7%)	1705 (-3.1%)	1752 (-6.7%)	10
Fan-out TTLS	Promise.all(100 steps)	1590 (-4.6%)	3569 (+6.8%)	3692 (+4.0%)	4409 (-34%) 💚	10
STSO	1020 steps (inline)	67 (-40%) 💚	133 (-10%)	155 (-11%)	230 (-35%) 💚	1019
WO	1020 steps	131497 (-12%)	131497 (-12%)	131497 (-12%)	131497 (-12%)	1
CRTT	first chunk (pooled)	88 (-9.3%)	120 (-20%) 💚	178 (+3.5%)	355 (+39%) 🔻	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	110 (-25%)	120 (-33%)	185 (-52%)	315 (-58%)	116 (-40%)	10
size sweep (100/s, 160B-12KB)	106 (-17%)	132 (-29%)	268 (-21%)	767 (+62%)	133 (-13%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	117 (-11%)	110 (-31%)	130 (-48%)	224 (-63%)	251 (-24%)	3
replay eve-gpt-5.6-sol-2000t (1x)	115 (-32%)	121 (-29%)	150 (-37%)	335 (-61%)	392 (-36%)	2
replay eve-gpt-5.6-sol-2000t (2x)	120 (-6%)	149 (-38%)	225 (-36%)	457 (-28%)	254 (-11%)	3

📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 148790ms → this run 131259ms (Δ -17531ms, -12%)

 50-100 ms ┃ main 0 this 3 +3
100-150 ms █████████████████████░░┃ main 779 this 872 +93
150-200 ms ███┃█ main 181 this 128 -53
200-250 ms ┃ main 31 this 9 -22
250-300 ms ┃ main 11 this 6 -5
300-350 ms ┃ main 6 this 0 -6
350-400 ms ┃ main 6 this 0 -6
400-450 ms ┃ main 1 this 1 +0
450-500 ms ┃ main 1 this 0 -1
500-550 ms ┃ main 1 this 0 -1
600-650 ms ┃ main 1 this 0 -1
650-700 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)

variant RTT 1ms→5s+ avg p50 p90 p99 n
control ·····▁█▇▁···· 103.2 (-32%) 96 (-24%) 185 (-52%) 315 (-58%) 3000
sweep ······▆█▁▁··· 126.4 (-17%) 107 (-15%) 268 (-21%) 767 (+62%) 3000
gw 1x ·····▁█▅▁···· 95.8 (-30%) 92 (-23%) 130 (-48%) 224 (-63%) 5295
eve 1x ·····▁█▇▁▁··· 103.6 (-35%) 94 (-27%) 150 (-37%) 335 (-61%) 5186
eve 2x ·····▁▅█▂▁··· 128.1 (-32%) 113 (-31%) 225 (-36%) 457 (-28%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▂▃▂▁▅▂▁▃▃ 95–124ms
sweep █▅▃▂▂▄▂▄▂▁ 105–169ms
gw 1x ▃█▃▄▂▂▂▁▂▄ 91–107ms
eve 1x ▃▂▃▃▁█▃▄▂▁ 86–146ms
eve 2x ▂▃▃▁▁▃▆█▅▂ 103–175ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep ▅█▆▃▂▁▂ 123–131ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▅▃▄█▃▅▅▁▃▆ 26–35ms
sweep ▅▆▇▄▇▄▆█▄▁ 41–56ms
gw 1x ▁█▃▃▃▄▆▅▄▅ 23–32ms
eve 1x ▂▃▁▁▂█▅▅▃▃ 19–29ms
eve 2x ▂▄▁▁▁▂█▄▂▆ 19–26ms

ℹ️ Metric definitions & methodology

_{Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.}

_{The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.}

_{The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.}

_{Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.}

_{Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)}

_{Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles}

_{Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600}

_{All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).}

_{Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.}

[github-actions]

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenario	outcome	events	virt	replay	violations
✅ smoke-no-steps	completed	3	0ms	ok	0
✅ smoke-one-step	completed	6	0ms	ok	0
✅ hook-at-step-started	completed	12	0ms	ok	0
✅ hook-at-step-completed	completed	12	0ms	ok	0
✅ hook-at-hook-created	completed	12	0ms	ok	0
✅ deadline-hook-wins	completed	7	1.0h	ok	0
✅ deadline-expires	completed	7	1.0h	ok	0
✅ long-sleep	completed	11	30.0d	ok	0
✅ hook-never-arrives	stalled	3	0ms	skipped	0
✅ step-retries-twice	completed	10	2.0s	ok	0
✅ parallel-steps	completed	9	0ms	ok	0
✅ hook-on-execution-state	completed	12	0ms	ok	0
✅ peek-hook-before-branch	completed	12	0ms	ok	0
✅ peek-hook-after-branch	completed	12	0ms	ok	0
✅ peek-hook-at-registration	completed	12	0ms	ok	0
✅ race-hook-before-probe	completed	12	0ms	ok	0
✅ race-hook-after-probe	completed	12	0ms	ok	0
✅ race-duplicate-delivery	completed	13	0ms	ok	0
✅ attr-hook-before-step	completed	11	0ms	ok	0
✅ attr-hook-after-step	completed	11	0ms	ok	0
✅ attr-from-step-body	completed	13	0ms	ok	0
✅ fork-hook-after-timeout	completed	14	1.0m	ok	0
✅ fork-hook-before-timeout	completed	14	1.0m	ok	0
✅ count-hook-after-timeout	completed	17	1.0m	ok	0
✅ count-hook-before-timeout	completed	20	1.0m	ok	0
✅ stale-read-step-count-fork	completed	20	1.0m	ok	0
✅ stale-read-equal-step-counts	completed	14	1.0m	ok	0
✅ step-vs-step-fork	completed	12	0ms	ok	0
✅ step-vs-step-fork-fenced	completed	12	0ms	ok	0
✅ fence-catches-benign-direction	completed	12	5ms	ok	0
✅ in-flight-before-decision	completed	17	1.0m	ok	0
❌ in-flight-before-decision-counted	completed	17	1.0m	ok	0
✅ in-flight-after-decision	completed	19	2.0m	ok	0
✅ stale-read-step-count-fork-fenced	completed	20	1.0m	ok	0
✅ fork-hook-wins	completed	13	1.0m	ok	0
✅ fork-timeout-wins	completed	13	1.0m	ok	0
✅ unclaimed-payload-under-fork	completed	17	1.0m	ok	0
✅ claimed-payload-under-fork	completed	17	1.0m	ok	0
✅ writers-independent-step-bodies	completed	12	0ms	ok	0
✅ writers-scripted-tempo	completed	12	0ms	ok	0
✅ cancel-mid-step	cancelled	7	0ms	skipped	0

Full trace: world-sim.txt

[vercel]

Additional Suggestion:

Test asserts SPEC_VERSION_CURRENT === SPEC_VERSION_SUPPORTS_COMPRESSION (5) but the PR bumped SPEC_VERSION_CURRENT to SPEC_VERSION_SUPPORTS_SLOT_IDENTITY (6), so the assertion evaluates 6 === 5 and fails.

[github-actions]

Framework	Cold replay	Step reg.	Framework output
hono	304.7 KiB (+103.8 KiB) ⚠️	40.6 KiB (±0)	2.58 MiB (+830.6 KiB)
nextjs-turbopack	305.0 KiB (+98.7 KiB) ⚠️	439 B (±0)	562.1 KiB (-203.4 KiB)

About these numbers

Sizes are gzip; parentheses show the change against main.
Cold replay and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.
⚠️ marks growth past the threshold. Add the allow-bundle-size-growth label to accept it.

770e721 · run