[core] Fix Run.returnValue retrying its accessor step for a terminal run by pranaygp · Pull Request #4326 · vercel/workflow

pranaygp · 2026-09-23T18:47:00Z

Closes #4288.

Run.prototype.returnValue is a built-in step, so whatever it throws is classified by the step executor's retry policy. When the target run is already terminal the accessor's read succeeds and it throws WorkflowRunFailedError (or WorkflowRunCancelledError for a cancelled run). Neither carried a fatal marker, so the executor treated them as transient and retried — even though a terminal run is immutable, so every retry re-reads the same record and throws the same error.

Impact

Driving the accessor through every attempt (the issue's repro runs only the first) shows two consequences:

the accessor body runs 4 times instead of 1, and the remote failure reaches the caller ~3s late (plus queue redelivery overhead in a real World), with three error-level log lines per await;
once the budget is spent the executor wraps the error, so what the caller's workflow finally catches is a FatalError whose message is Step "…" failed after 3 retries: …. WorkflowRunFailedError.is(err) is false there, and the error is only reachable via err.cause — which contradicts what the WorkflowRunFailedError docs tell users to write. The same await outside a workflow (not a step) surfaces WorkflowRunFailedError directly, so the two contexts disagree today.

Fix

Mark WorkflowRunFailedError and WorkflowRunCancelledError non-retryable (fatal = true, the own-property opt-in FatalError.is() already recognises). Both are thrown only after a run has been read successfully and found terminal, so there is no case where retrying them can produce a different answer. Errors from failing to read the run — transport blips, WorkflowRunNotFoundError during a resilient start — are a different case and stay retryable; a test pins that.

WorkflowRunCancelledError is included because it is the same accessor reaching the same conclusion about the same immutable state; leaving it out would fix half of one code path.

Nothing else keys off these two classes: classifyRunError doesn't consult FatalError, so run errorCode is unchanged (USER_ERROR either way), and the only other construction site is the onRunFailed lifecycle payload, which is not thrown for retry classification.

Tests

packages/errors/src/terminal-run-error.test.ts — pins the marker on both classes and its absence on WorkflowRunNotCompletedError / WorkflowRunNotFoundError.
packages/core/src/runtime/run-return-value-terminal-retry.test.ts — the end-to-end repro: seeds a terminal run in a real world-local, registers the actual Run.prototype.returnValue getter as a step, runs it through the real executeStep, and asserts the step emits step_created, step_started, step_failed with no step_retrying, carrying an intact WorkflowRunFailedError. A third case pins that a completed run still resolves normally.

Both failed on main before the fix (expected 'retry' to be 'failed').

Out of scope, noted while here

runId and errorCode do not survive a step boundary: the generic Error reducer carries only name/message/stack/cause, so no SDK error class without a dedicated reducer keeps its extra fields. That is orthogonal to the retry decision and affects every such class, so this PR identifies the run through the error message instead of widening into the serializer.

Docs Preview

Page	Preview
`WorkflowRunFailedError`	https://workflow-docs-git-fix-returnvalue-retry.vercel.sh/docs/api-reference/workflow-errors/workflow-run-failed-error
`WorkflowRunCancelledError`	https://workflow-docs-git-fix-returnvalue-retry.vercel.sh/docs/api-reference/workflow-errors/workflow-run-cancelled-error

(Behind deployment protection; requires Vercel team access.)

🤖 Generated with Claude Code

…al run The accessor is a built-in step, so what it throws is subject to the step executor's retry policy. A terminal run is immutable: once the accessor has successfully read one, re-running the body re-reads the same record. Today `WorkflowRunFailedError` / `WorkflowRunCancelledError` carry no fatal marker, so the executor spends the whole retry budget before the failure reaches the caller — and then replaces it with its retry-exhaustion `FatalError` wrapper, so `WorkflowRunFailedError.is(err)` is false in the caller's workflow. Closes #4288. Co-Authored-By: Pranay Prakash <1797812+pranaygp@users.noreply.github.com>

vercel · 2026-09-23T18:47:03Z

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

Project	Deployment	Actions	Updated
example-nextjs-workflow-turbopack	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
example-nextjs-workflow-webpack	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
example-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-astro-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-express-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-fastify-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-hono-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-nestjs-workflow	Building	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-nitro-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-nuxt-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-python-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-sveltekit-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-tanstack-start-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workbench-vite-workflow	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workflow-docs	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workflow-swc-playground	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workflow-tarballs	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC
workflow-web	Ready	Preview, v0	Sep 23, 2026 7:13pm UTC

changeset-bot · 2026-09-23T18:47:06Z

🦋 Changeset detected

Latest commit: 3730f7e

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

This PR includes changesets to release 20 packages

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

github-actions · 2026-09-23T18:47:21Z

🧪 E2E Test Results

✅ All tests passed

⚠️ 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.

onRunFailed receives the hydrated error with errorCode and cause (nextjs-webpack · local-dev / local / node / stable)
plainModuleDoneHook resumed via plain API route (o2flow shape) (nextjs-webpack · local-dev / local / node / stable)
promiseAllWorkflow via pages router (nextjs-webpack · local-dev / local / node / stable)
sleepingWorkflow via pages router (nextjs-webpack · local-dev / local / node / stable)

🛠 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.

cold-start-warmup · suite warmup (tanstack-start) · at 19:51:17Z · abandoned wrun_01M37X3SE1QFB93S7GFRKV6AX9 · (+1 more)
run-pickup-stall · hookCleanupTestWorkflow - hook token reuse after workflow completion (nextjs-webpack) · at 19:56:55Z · abandoned wrun_01M37XEYH21VMJ3785JJ276KCV

E2E Test Summary

Summary

	Passed	Skipped	Total
✅ ▲ Vercel Production	3670	731	4401
✅ 💻 Local Development	4014	550	4564
✅ 📦 Local Production	4014	550	4564
✅ 🐘 Local Postgres	4014	550	4564
✅ 🪟 Windows	324	2	326
✅ 🌐 Cross-language Conformance	68	76	144
✅ vercel-http-transport	825	153	978
✅ vercel-multi-region	27	0	27
✅ vercel-ws-transport	559	93	652
Total	17515	2705	20220

Details by Category

✅ ▲ Vercel Production

App	Passed	Skipped
✅ astro-node	133	30
✅ astro-quickjs	133	30
✅ example-node	133	30
✅ example-quickjs	133	30
✅ express-node	133	30
✅ express-quickjs	133	30
✅ fastify-node	133	30
✅ fastify-quickjs	133	30
✅ hono-node	133	30
✅ hono-quickjs	133	30
✅ nest-node	133	30
✅ nest-quickjs	133	30
✅ nextjs-turbopack-node	160	3
✅ nextjs-turbopack-quickjs	160	3
✅ nextjs-webpack-node	160	3
✅ nextjs-webpack-quickjs	160	3
✅ nitro-node	133	30
✅ nitro-quickjs	133	30
✅ nuxt-node	133	30
✅ nuxt-quickjs	133	30
✅ python-node	66	97
✅ sveltekit-node	152	11
✅ sveltekit-quickjs	152	11
✅ tanstack-start-node	133	30
✅ tanstack-start-quickjs	133	30
✅ vite-node	133	30
✅ vite-quickjs	133	30

✅ 💻 Local Development

App	Passed	Skipped
✅ astro-stable-node	134	29
✅ astro-stable-quickjs	134	29
✅ express-stable-node	134	29
✅ express-stable-quickjs	134	29
✅ fastify-stable-node	134	29
✅ fastify-stable-quickjs	134	29
✅ hono-stable-node	134	29
✅ hono-stable-quickjs	134	29
✅ nest-stable-node	134	29
✅ nest-stable-quickjs	134	29
✅ nextjs-turbopack-canary-node	162	1
✅ nextjs-turbopack-canary-quickjs	162	1
✅ nextjs-turbopack-stable-node	162	1
✅ nextjs-turbopack-stable-quickjs	162	1
✅ nextjs-webpack-canary-node	162	1
✅ nextjs-webpack-canary-quickjs	162	1
✅ nextjs-webpack-stable-node	162	1
✅ nextjs-webpack-stable-quickjs	162	1
✅ nitro-stable-node	134	29
✅ nitro-stable-quickjs	134	29
✅ nuxt-stable-node	134	29
✅ nuxt-stable-quickjs	134	29
✅ sveltekit-stable-node	153	10
✅ sveltekit-stable-quickjs	153	10
✅ tanstack-start-node	134	29
✅ tanstack-start-quickjs	134	29
✅ vite-stable-node	134	29
✅ vite-stable-quickjs	134	29

✅ 📦 Local Production

App	Passed	Skipped
✅ astro-stable-node	134	29
✅ astro-stable-quickjs	134	29
✅ express-stable-node	134	29
✅ express-stable-quickjs	134	29
✅ fastify-stable-node	134	29
✅ fastify-stable-quickjs	134	29
✅ hono-stable-node	134	29
✅ hono-stable-quickjs	134	29
✅ nest-stable-node	134	29
✅ nest-stable-quickjs	134	29
✅ nextjs-turbopack-canary-node	162	1
✅ nextjs-turbopack-canary-quickjs	162	1
✅ nextjs-turbopack-stable-node	162	1
✅ nextjs-turbopack-stable-quickjs	162	1
✅ nextjs-webpack-canary-node	162	1
✅ nextjs-webpack-canary-quickjs	162	1
✅ nextjs-webpack-stable-node	162	1
✅ nextjs-webpack-stable-quickjs	162	1
✅ nitro-stable-node	134	29
✅ nitro-stable-quickjs	134	29
✅ nuxt-stable-node	134	29
✅ nuxt-stable-quickjs	134	29
✅ sveltekit-stable-node	153	10
✅ sveltekit-stable-quickjs	153	10
✅ tanstack-start-node	134	29
✅ tanstack-start-quickjs	134	29
✅ vite-stable-node	134	29
✅ vite-stable-quickjs	134	29

✅ 🐘 Local Postgres

App	Passed	Skipped
✅ astro-stable-node	134	29
✅ astro-stable-quickjs	134	29
✅ express-stable-node	134	29
✅ express-stable-quickjs	134	29
✅ fastify-stable-node	134	29
✅ fastify-stable-quickjs	134	29
✅ hono-stable-node	134	29
✅ hono-stable-quickjs	134	29
✅ nest-stable-node	134	29
✅ nest-stable-quickjs	134	29
✅ nextjs-turbopack-canary-node	162	1
✅ nextjs-turbopack-canary-quickjs	162	1
✅ nextjs-turbopack-stable-node	162	1
✅ nextjs-turbopack-stable-quickjs	162	1
✅ nextjs-webpack-canary-node	162	1
✅ nextjs-webpack-canary-quickjs	162	1
✅ nextjs-webpack-stable-node	162	1
✅ nextjs-webpack-stable-quickjs	162	1
✅ nitro-stable-node	134	29
✅ nitro-stable-quickjs	134	29
✅ nuxt-stable-node	134	29
✅ nuxt-stable-quickjs	134	29
✅ sveltekit-stable-node	153	10
✅ sveltekit-stable-quickjs	153	10
✅ tanstack-start-node	134	29
✅ tanstack-start-quickjs	134	29
✅ vite-stable-node	134	29
✅ vite-stable-quickjs	134	29

✅ 🪟 Windows

App	Passed	Failed	Skipped
✅ nextjs-turbopack-node	162	0	1
✅ nextjs-turbopack-quickjs	162	0	1

✅ 🌐 Cross-language Conformance

App	Passed	Failed	Skipped
✅ python	68	0	76

✅ vercel-http-transport

App	Passed	Skipped
✅ example	133	30
✅ express	133	30
✅ hono	133	30
✅ nextjs-turbopack	160	3
✅ nitro	133	30
✅ vite	133	30

✅ vercel-multi-region

App	Passed	Failed	Skipped
✅ nextjs-turbopack	27	0	0

✅ vercel-ws-transport

App	Passed	Skipped
✅ example	133	30
✅ express	133	30
✅ nextjs-turbopack	160	3
✅ vite	133	30

📋 View full workflow run

github-actions · 2026-09-23T18:47:22Z

📊 Workflow Benchmarks

commit 3730f7e · Wed, 23 Sep 2026 20:14:32 GMT · run logs

Backend: vercel · app: nextjs-turbopack

Metric	Scenario	Best (ms)	P75 (ms)	P90 (ms)	P99 (ms)	Samples
TTFS	step	2391 (+72%) 🔻	2562 🔴 (+47%) 🔻	2583 🔴 (+45%) 🔻	2610 🔴 (+14%)	30
TTFS	stream	208 (+28%) 🔻	2523 🔴 (+48%) 🔻	2563 🔴 (+47%) 🔻	2624 🔴 (+34%) 🔻	30
TTFS	hook + stream	2711 (+330%) 🔻	2935 🔴 (+41%) 🔻	2991 🔴 (+32%) 🔻	3314 🔴 (+25%) 🔻	30
Fan-out TTFS	Promise.all(100 steps)	570 (+16%) 🔻	898 (+51%) 🔻	1142 (+76%) 🔻	3715 (+64%) 🔻	10
Fan-out TTLS	Promise.all(100 steps)	2299 (+26%) 🔻	3888 (+40%) 🔻	4061 (+41%) 🔻	13050 (+51%) 🔻	10
STSO	1020 steps (inline)	101 (-11%)	140 (+1.4%)	157 (+3.3%)	205 (-82%) 💚	1019
WO	1020 steps	141727 (-10%)	141727 (-10%)	141727 (-10%)	141727 (-10%)	1
CRTT	first chunk (pooled)	76 (+23%) 🔻	122 (+20%) 🔻	143 (-26%) 💚	208 (-18%) 💚	28

Streams

Scenario	CRTT 1st	p75	p90	p99	CDV max	iters
paced control (100/s, 60B)	123 (+50%)	200 (+14%)	258 (-25%)	396 (-52%)	174 (+24%)	10
size sweep (100/s, 160B-12KB)	95.5 (+19%)	200 (-18%)	326 (-13%)	570 (-12%)	154 (-28%)	10
replay gateway-gpt-5.4-nano-2000t (1x)	99 (+8%)	143 (-27%)	172 (-93%)	338 (-92%)	200 (-38%)	3
replay eve-gpt-5.6-sol-2000t (1x)	84 (-52%)	143 (-23%)	180 (-33%)	310 (-90%)	271 (-87%)	2
replay eve-gpt-5.6-sol-2000t (2x)	89 (-6%)	261 (-29%)	338 (-26%)	593 (-3%)	277 (-25%)	3

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

1020 steps (inline)

Cumulative STSO time: main 157647ms → this run 141541ms (Δ -16106ms, -10%)

  100-150 ms  ██████████████████████┃█  main 899  this 879   -20
  150-200 ms  ██┃                       main  69  this 127   +58
  200-250 ms  ┃                         main  13  this   6    -7
  250-300 ms  ┃                         main   7  this   1    -6
  300-350 ms  ┃                         main   2  this   2    +0
  350-400 ms  ┃                         main   2  this   1    -1
  400-450 ms  ┃                         main   1  this   3    +2
  500-550 ms  ┃                         main   1  this   0    -1
  550-600 ms  ┃                         main   5  this   0    -5
  700-750 ms  ┃                         main   2  this   0    -2
  750-800 ms  ┃                         main   2  this   0    -2
  800-850 ms  ┃                         main   2  this   0    -2
  850-900 ms  ┃                         main   1  this   0    -1
  900-950 ms  ┃                         main   1  this   0    -1
1050-1100 ms  ┃                         main   1  this   0    -1
1100-1150 ms  ┃                         main   1  this   0    -1
1150-1200 ms  ┃                         main   7  this   0    -7
1200-1250 ms  ┃                         main   1  this   0    -1
1300-1350 ms  ┃                         main   1  this   0    -1
1450-1500 ms  ┃                         main   1  this   0    -1

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

variant  RTT 1ms→5s+             avg         p50         p90         p99     n
control  ······▂█▃····     158 (+5%)  145 (+21%)  258 (-25%)  396 (-52%)  3000
sweep    ······▂█▃▁···    159 (-10%)  138 (-12%)  326 (-13%)  570 (-12%)  3000
gw 1x    ·····▁▃█▁····  124.7 (-53%)  120 (-10%)  172 (-93%)  338 (-92%)  5295
eve 1x   ·····▁▅█▁▁···  120.5 (-33%)  104 (-25%)  180 (-33%)  310 (-90%)  5186
eve 2x   ·····▁▂█▅▁···  187.9 (-13%)   151 (-7%)  338 (-26%)   593 (-3%)  7779

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

control  ▇█▆▂▃▂▃▂▂▁  137–196ms
sweep    █▆▅▄▃▁▇█▆▃  136–175ms
gw 1x    █▄▃▄▃▃█▁▃▂  115–140ms
eve 1x   ▄▁▅▁▁█▆▆▃▄  105–142ms
eve 2x   ▂▁▃▁▁▂▃█▄▂  149–308ms

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

sweep  ▁▆█▇█▇▁  156–161ms

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

control  ▇█▃▃▁▃▄▃▂▁  42–73ms
sweep    ▃▁█▇▅▄▆▃▅▁  52–70ms
gw 1x    █▃▄▁▃▂▆▃▂▃  35–48ms
eve 1x   ▄▄▅▁▄█▅▇▃▆  21–27ms
eve 2x   █▄▁▁▃▄▄▂▄▃  25–36ms

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-2000t eaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t 6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041}

_{🔴 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 · 2026-09-23T18:51:03Z

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 42 total

fence=per-spec

scenario	outcome	events	virt	replay
✅ `smoke-no-steps`	completed	3	0ms	ok
✅ `smoke-one-step`	completed	6	0ms	ok
✅ `hook-at-step-started`	completed	12	0ms	ok
✅ `hook-at-step-completed`	completed	12	0ms	ok
✅ `hook-at-hook-created`	completed	12	0ms	ok
✅ `deadline-hook-wins`	completed	7	1.0h	ok
✅ `deadline-expires`	completed	7	1.0h	ok
✅ `step-vs-timer-early-settlement`	completed	8	1.0h	ok
✅ `long-sleep`	completed	11	30.0d	ok
✅ `hook-never-arrives`	stalled	3	0ms	skipped
✅ `step-retries-twice`	completed	10	2.0s	ok
✅ `parallel-steps`	completed	9	0ms	ok
✅ `hook-on-execution-state`	completed	12	0ms	ok
✅ `peek-hook-before-branch`	completed	12	0ms	ok
✅ `peek-hook-after-branch`	completed	12	0ms	ok
✅ `peek-hook-at-registration`	completed	12	0ms	ok
✅ `race-hook-before-probe`	completed	12	0ms	ok
✅ `race-hook-after-probe`	completed	12	0ms	ok
✅ `race-duplicate-delivery`	completed	13	0ms	ok
✅ `attr-hook-before-step`	completed	11	0ms	ok
✅ `attr-hook-after-step`	completed	11	0ms	ok
✅ `attr-from-step-body`	completed	13	0ms	ok
✅ `fork-hook-after-timeout`	completed	14	1.0m	ok
✅ `fork-hook-before-timeout`	completed	14	1.0m	ok
✅ `count-hook-after-timeout`	completed	17	1.0m	ok
✅ `count-hook-before-timeout`	completed	20	1.0m	ok
✅ `stale-read-step-count-fork`	completed	20	1.0m	ok
✅ `stale-read-equal-step-counts`	completed	14	1.0m	ok
✅ `step-vs-step-fork`	completed	12	0ms	ok
✅ `step-vs-step-fork-fenced`	completed	12	0ms	ok
✅ `fence-catches-benign-direction`	completed	12	5ms	ok
✅ `in-flight-before-decision`	completed	17	1.0m	ok
❌ `in-flight-before-decision-counted`	completed	17	1.0m	ok
✅ `in-flight-after-decision`	completed	19	2.0m	ok
✅ `stale-read-step-count-fork-fenced`	completed	20	1.0m	ok
✅ `fork-hook-wins`	completed	13	1.0m	ok
✅ `fork-timeout-wins`	completed	13	1.0m	ok
✅ `unclaimed-payload-under-fork`	completed	17	1.0m	ok
✅ `claimed-payload-under-fork`	completed	17	1.0m	ok
✅ `writers-independent-step-bodies`	completed	12	0ms	ok
✅ `writers-scripted-tempo`	completed	12	0ms	ok
✅ `cancel-mid-step`	cancelled	7	0ms	skipped

Full trace: world-sim.txt

github-actions · 2026-09-23T18:51:51Z

Framework	Flow route	Step reg.	Framework output
hono	251.1 KiB (±0)	93.8 KiB (+309 B)	1.90 MiB (-565 B)
nextjs-turbopack	258.5 KiB (±0)	426 B (±0)	903.1 KiB (-353 B)

About these numbers

Sizes are gzip; parentheses show the change against main.
Flow route and Step reg. gate this job, on raw bytes rather than the gzip shown, at max(2%, 50.0 KiB). Framework output is informational.

3730f7e · run

`WorkflowRunFailedError` and `WorkflowRunCancelledError` are only thrown after a run has been read *successfully* and found in a terminal state. A terminal run is immutable, so re-running the read returns the same record and throws the same error. Neither carried a fatal marker, so `Run.returnValue` — a built-in step — spent its whole retry budget doing exactly that when a parent awaited a child that had already failed, and the error the caller finally caught was the executor's retry-exhaustion `FatalError` wrapper rather than the `WorkflowRunFailedError` the docs point it at. Mark both non-retryable. Errors from *failing to read* the run (transport blips, `WorkflowRunNotFoundError` during a resilient start) are a different case and stay retryable. Closes #4288. Co-Authored-By: Pranay Prakash <1797812+pranaygp@users.noreply.github.com>

Copilot

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

kvnloo · 2026-09-23T23:48:30Z

one edge case i think is worth pinning: the first accessor step now correctly sees the terminal-run error as non-retryable, but does that property survive the step boundary?

from what i can tell, "WorkflowRunFailedError" falls through the generic error serializer, which keeps the name/message/stack/cause but drops custom fields like "fatal".

so after the workflow catches the accessor failure, "WorkflowRunFailedError.is(error)" is still true, but "FatalError.is(error)" looks like it would be false.

that seems observable if the workflow passes that error into another step and that step rethrows it — we'd start retrying the same terminal error again.

would it be worth adding a round-trip assertion to the existing regression that "FatalError.is(error)" is still true after "failureOf()" hydrates it? that should tell us whether this needs serializer support or just narrower docs.

…run (#4326) Signed-off-by: Pranay Prakash <pranay.gp@gmail.com>

github-actions · 2026-09-24T00:17:39Z

Backport PR opened against stable: #4346. Merge conflicts were resolved by AI — please review carefully. (backport job run)

Surface `createHook({ experimental_force: true })` (#4193) on the hook conflict, idempotency, and event-sourcing pages, and note on the child workflows section that a failed or canceled child's `returnValue` throws on the first attempt without retrying (#4326). Co-Authored-By: Claude <noreply@anthropic.com> Co-Authored-By: Pranay Prakash <1797812+pranaygp@users.noreply.github.com>

* [docs] Cross-link hook takeover and terminal child-run errors Surface `createHook({ experimental_force: true })` (#4193) on the hook conflict, idempotency, and event-sourcing pages, and note on the child workflows section that a failed or canceled child's `returnValue` throws on the first attempt without retrying (#4326). Co-Authored-By: Claude <noreply@anthropic.com> Co-Authored-By: Pranay Prakash <1797812+pranaygp@users.noreply.github.com> * [docs] Address review: use experimental_force for superseding the owner Co-Authored-By: Claude <noreply@anthropic.com> Co-Authored-By: Pranay Prakash <1797812+pranaygp@users.noreply.github.com> --------- Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com> Co-authored-by: Pranay Prakash <1797812+pranaygp@users.noreply.github.com>

…run (#4326) (#4346)

vercel Bot deployed to Preview – workbench-python-workflow September 23, 2026 18:47 View deployment

vercel Bot deployed to Preview – workflow-web September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – workbench-tanstack-start-workflow September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – workbench-hono-workflow September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – workbench-nestjs-workflow September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – workbench-fastify-workflow September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – workbench-nitro-workflow September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – example-workflow September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – workflow-tarballs September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – workbench-express-workflow September 23, 2026 18:48 View deployment

vercel Bot deployed to Preview – workbench-python-workflow September 23, 2026 19:09 View deployment

vercel Bot deployed to Preview – workflow-web September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workbench-hono-workflow September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workbench-nitro-workflow September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workbench-express-workflow September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workbench-nestjs-workflow September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workflow-tarballs September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – example-workflow September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – example-nextjs-workflow-webpack September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workbench-fastify-workflow September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workbench-vite-workflow September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workbench-astro-workflow September 23, 2026 19:10 View deployment

vercel Bot deployed to Preview – workflow-docs September 23, 2026 19:11 View deployment

vercel Bot deployed to Preview – workflow-swc-playground September 23, 2026 19:13 View deployment

pranaygp closed this Sep 23, 2026

pranaygp reopened this Sep 23, 2026

pranaygp closed this Sep 23, 2026

pranaygp reopened this Sep 23, 2026

pranaygp marked this pull request as ready for review September 23, 2026 20:15

pranaygp requested a review from a team as a code owner September 23, 2026 20:15

Copilot AI lite review requested due to automatic review settings September 23, 2026 20:15

Copilot AI reviewed Sep 23, 2026

pranaygp mentioned this pull request Sep 23, 2026

Run.returnValue retries its accessor step for an already-failed run #4288

Closed

VaguelySerious approved these changes Sep 24, 2026

View reviewed changes

VaguelySerious merged commit 2694663 into main Sep 24, 2026
498 of 509 checks passed

VaguelySerious deleted the fix-returnvalue-retry branch September 24, 2026 00:08

github-actions Bot mentioned this pull request Sep 24, 2026

Version Packages (beta) #4345

Merged

github-actions Bot added a commit that referenced this pull request Sep 24, 2026

[core] Fix Run.returnValue retrying its accessor step for a terminal …

7d226ff

…run (#4326) Signed-off-by: Pranay Prakash <pranay.gp@gmail.com>

This was referenced Sep 24, 2026

Backport #4326: [core] Fix Run.returnValue retrying its accessor step for a terminal run #4346

Merged

Version Packages (beta) Jokacar10/jkworkflow#548

Open

Version Packages (beta) ben-vargas/workflow#104

Open

pranaygp mentioned this pull request Sep 24, 2026

[docs] Cross-link hook takeover and terminal child-run errors #4353

Merged

github-actions Bot mentioned this pull request Sep 24, 2026

Version Packages Thegreatsura/workflow#1

Open

pranaygp mentioned this pull request Sep 24, 2026

[docs] Document hookForceClaim for World authors and spec version 8 #4361

Merged

VaguelySerious pushed a commit that referenced this pull request Sep 24, 2026

[core] Fix Run.returnValue retrying its accessor step for a terminal …

5e6c79c

…run (#4326) (#4346)

This was referenced Sep 24, 2026

Version Packages #4182

Merged

Version Packages (beta) azaj01/workflow#1

Open

Version Packages (beta) erickirt/workflow#523

Open

Navigation Menu

[core] Fix Run.returnValue retrying its accessor step for a terminal run - #4326

[core] Fix Run.returnValue retrying its accessor step for a terminal run#4326
VaguelySerious merged 2 commits into
mainfrom
fix-returnvalue-retry

pranaygp commented Sep 23, 2026 •

edited

Loading

Uh oh!

vercel Bot commented Sep 23, 2026 •

edited

Loading

Uh oh!

changeset-bot Bot commented Sep 23, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 23, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 23, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 23, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 23, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

kvnloo commented Sep 23, 2026

Uh oh!

Uh oh!

github-actions Bot commented Sep 24, 2026

Uh oh!

This branch was successfully deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

pranaygp commented Sep 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Impact

Fix

Tests

Out of scope, noted while here

Docs Preview

Uh oh!

vercel Bot commented Sep 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

changeset-bot Bot commented Sep 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🦋 Changeset detected

Uh oh!

github-actions Bot commented Sep 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🧪 E2E Test Results

⚠️ Flaky E2E Tests (passed on retry)

🛠 Infra Events (absorbed by the harness)

E2E Test Summary

Uh oh!

github-actions Bot commented Sep 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

📊 Workflow Benchmarks

Uh oh!

github-actions Bot commented Sep 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Sim World

Uh oh!

github-actions Bot commented Sep 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

kvnloo commented Sep 23, 2026

Uh oh!

Uh oh!

github-actions Bot commented Sep 24, 2026

Uh oh!

This branch was successfully deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

pranaygp commented Sep 23, 2026 •

edited

Loading

vercel Bot commented Sep 23, 2026 •

edited

Loading

changeset-bot Bot commented Sep 23, 2026 •

edited

Loading

github-actions Bot commented Sep 23, 2026 •

edited

Loading

github-actions Bot commented Sep 23, 2026 •

edited

Loading

github-actions Bot commented Sep 23, 2026 •

edited

Loading

github-actions Bot commented Sep 23, 2026 •

edited

Loading