test(eve): reproduce stale input after steered cancellation by e0-gh-vercel-connect[bot] · Pull Request #3459 · vercel/eve

e0-gh-vercel-connect · 2026-09-17T07:31:41Z

Regression test for #3458

Steering a background subagent while it is parked on a proxied approval can leave a pending input batch in the child after cancellation. The retired parent task can no longer route the approval response back to that child, so later turns stop after one model cycle and never notify the parent.

Why

The cancellation epilogue clears proxy routes but preserves ordinary pending input batches. The focused reproduction confirms that this state is treated as a live input wait by model-call batching even though no route can answer it.

Approach

The new integration regression seeds a child session with a proxied tool-approval batch, runs the production cancellation step, and checks both invariants that should hold after cancellation:

the child has no unanswerable pending input batch;
a later turn can use its full model-call budget.

This draft intentionally contains only the failing regression test. The test fails on origin/main before any production change.

Validation

pnpm --filter eve exec vitest run --config vitest.integration.config.ts src/execution/issue-3458.repro.integration.test.ts

Result: fails with both expected failures on origin/main at 4c3b93356aa1b60672fad17c4b686c8ae66e0d23.

Relevant issue: #3458

vercel · 2026-09-17T07:31:47Z

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

Project	Deployment	Actions	Updated
eve-docs	Ready	Preview, v0	Sep 17, 2026 7:41am UTC
eve-pkg	Ready	Preview, v0	Sep 17, 2026 7:41am UTC

github-actions · 2026-09-17T07:35:59Z

Bundle + Package Summary: `apps/fixtures/weather-agent`

Key takeaways

No notable deltas vs main (856d765).

Delta vs `main (856d765)`

Area	Metric	Baseline	Current	Delta
Package	Packed tarball	8.76 MB	8.76 MB	+862 B ⚠️
Package	Unpacked publish size	32.85 MB	32.85 MB	+147 B ⚠️
Package	Installed footprint	78.92 MB	78.92 MB	+147 B ⚠️
Package	Published files	4030	4030	0
Package	Installed files	7990	7990	0
Package	Installed package instances	33	33	0
Package	Distinct installed package names	32	32	0
Package	Installed dependency edges	51	51	0
Package	Installed optional peer edges	8	8	0
Runtime	Unique function payloads	2	2	0
Runtime	Total function bytes	20.41 MB	20.41 MB	-308 B ✅
Runtime	Public routes	18	18	0

Changed function payloads vs main (856d765) (2)

Function	Status	Baseline	Current	Delta	Route changes
`functions/__server.func`	changed	10.21 MB	10.21 MB	-154 B ✅	none
`functions/.well-known/workflow/v1/flow.func`	changed	10.21 MB	10.21 MB	-154 B ✅	none

`eve init` install

Metric	Baseline	Current	Delta
Installed footprint	117.14 MB	117.14 MB	+147 B ⚠️
Installed packages	113	113	0
dependencies	4	4	0
devDependencies	2	2	0
Dependency package bytes	49.54 MB	49.54 MB	+147 B ⚠️
devDependency package bytes	5.04 MB	5.04 MB	0 B ➖

Build Metadata

Preset: vercel
Nitro: nitro@3.0.260903-beta
Output directory: apps/fixtures/weather-agent/.vercel/output
Build metadata timestamp: 2026-09-17T07:42:20.627Z
Route aliases: 18 public, 1 internal (19 total aliases)
Vercel routes in config: 21
Severity legend: 🔴 dominant/large, 🟠 notable, 🟡 watch, ⚪ small

Package Drill-Down

Package Details

Package: eve@0.58.1
Package directory: packages/eve
Tarball: 8.76 MB (eve-0.58.1.tgz)
Unpacked payload: 32.85 MB across 4030 published files
Installed footprint: 78.92 MB across 7990 installed files
Installed root package: 31.42 MB
Installed dependencies: 47.50 MB
Installed package instances: 33
Distinct installed package names: 32
Installed dependency edges: 51
Installed optional peer edges: 8
Runtime dependencies: 2
Peer dependencies: 6 (5 optional)

Installed footprint is measured from an isolated temporary npm install of the packed tarball.
Graph metrics read only package.json files in package directories directly beneath a node_modules boundary, including nested boundaries. Each directory is one package instance; distinct names come from those manifests. Dependency edges count each unique name in dependencies or optionalDependencies per instance; optional peer edges count peerDependencies marked optional.

Heavy installed dependencies

eve: 31.42 MB (39.8%)
@rolldown/binding-linux-x64-gnu: 19.10 MB (24.2%)
ai: 7.56 MB (9.6%)
zod: 6.76 MB (8.6%)
undici: 3.51 MB (4.5%)

Publish payload breakdown

Published file size
🔴 dist/src/compiled/shadcn-registry/index.js       [#############...........] 9.76 MB 29.7%
🟠 dist/src/compiled/@photon-ai/chat-adapter-ime... [###.....................] 2.27 MB 6.9%
🟠 dist/src/compiled/@ai-sdk/code-mode/index.js     [#.......................] 1.03 MB 3.1%
🟡 dist/src/compiled/@vercel/blob/index.js          [#.......................] 901.4 kB 2.7%
🟡 dist/src/compiled/_chunks/workflow/signal-exi... [#.......................] 514.6 kB 1.6%
🔴 Other published files                            [########################] 18.38 MB 55.9%

Installed footprint breakdown

Installed package size
🔴 eve                             [########################] 31.42 MB 39.8%
🔴 @rolldown/binding-linux-x64-gnu [###############.........] 19.10 MB 24.2%
🔴 ai                              [######..................] 7.56 MB 9.6%
🔴 zod                             [#####...................] 6.76 MB 8.6%
🟠 undici                          [###.....................] 3.51 MB 4.5%
🟠 nitro                           [#.......................] 1.89 MB 2.4%
🔴 Other installed packages        [#######.................] 8.67 MB 11.0%

Runtime dependencies (2)

Package	Range	Notes
`nitro`	`3.0.260903-beta`
`undici`	`8.9.0`

Peer dependencies (6)

Package	Range	Notes
`@opentelemetry/api`	`^1.0.0`	optional peer
`ai`	`catalog:`
`braintrust`	`^3.0.0`	optional peer
`dd-trace`	`^6.13.0`	optional peer
`just-bash`	`^3.1.0`	optional peer
`microsandbox`	`^0.5.0`	optional peer

eve init install drill-down

`eve init` install details

Command: eve init my-agent
Package manager: npm
Installed footprint: 117.14 MB across 9884 installed files
Installed packages: 113 total (107 transitive-only)
dependencies: 4 direct packages totaling 49.54 MB
devDependencies: 2 direct packages totaling 5.04 MB
Other transitive package files: 62.55 MB

Installed footprint is measured from an isolated temporary eve init my-agent using the current packed eve tarball.

Heavy installed dependencies

eve: 31.42 MB (26.8%)
@typescript/typescript-linux-x64: 27.95 MB (23.9%)
@rolldown/binding-linux-x64-gnu: 19.10 MB (16.3%)
zod: 10.37 MB (8.9%)
ai: 7.56 MB (6.5%)

Installed footprint breakdown

Installed package size
🔴 eve                              [########################] 31.42 MB 26.8%
🔴 @typescript/typescript-linux-x64 [#####################...] 27.95 MB 23.9%
🔴 @rolldown/binding-linux-x64-gnu  [###############.........] 19.10 MB 16.3%
🔴 zod                              [########................] 10.37 MB 8.9%
🔴 ai                               [######..................] 7.56 MB 6.5%
🟠 undici                           [###.....................] 3.51 MB 3.0%
🔴 Other installed packages         [#############...........] 17.22 MB 14.7%

dependencies (4)

Package	Range	Installed size	Share
`@vercel/connect`	`2.2.0`	188.7 kB	0.2%
`ai`	`^7.0.105`	7.56 MB	6.5%
`eve`	`file:eve-0.58.1.tgz`	31.42 MB	26.8%
`zod`	`4.5.4`	10.37 MB	8.9%

devDependencies (2)

Package	Range	Installed size	Share
`@types/node`	`24.x`	2.54 MB	2.2%
`typescript`	`7.0.2`	2.50 MB	2.1%

Function Drill-Down

Payload Size Graph

Unique function payload size and share of total
🔴 functions/.well-known/workflow/v1/flow.func     [########################] 10.21 MB 50.0%
🔴 functions/__server.func                         [########################] 10.21 MB 50.0%

Top Function Payloads

🟠 functions/.well-known/workflow/v1/flow.func • 1 public route • 10.21 MB

Metric	Value
Public routes	`/.well-known/workflow/v1/flow`
Runtime	`nodejs24.x`
Handler	`index.mjs`
Payload	10.21 MB
Function files	10.21 MB across 116 files
Traced dependencies	0 B
Signal	🟠 Bundled file `index.mjs` is 2.07 MB (20.3%)

🟠 🔎 Dependency Analysis

📦 Bundled files:

Bundled file size
🟠 index.mjs                        [##########..............] 2.07 MB 20.3%
🟡 _libs/undici.mjs                 [#####...................] 980.8 kB 9.6%
🟡 _chunks/esm-Fqlolk7e.mjs         [###.....................] 723.3 kB 7.1%
🟡 _chunks/chatgpt-model.mjs        [###.....................] 698.7 kB 6.8%
🟡 _chunks/signal-exit-Dsy-TT0V.mjs [###.....................] 616.2 kB 6.0%
🔴 Other bundled files              [########################] 5.11 MB 50.1%

🧾 Vercel Config

{
  "handler": "index.mjs",
  "launcherType": "Nodejs",
  "shouldAddHelpers": false,
  "supportsResponseStreaming": true,
  "runtime": "nodejs24.x",
  "maxDuration": "max",
  "experimentalTriggers": [
    {
      "type": "queue/v2beta",
      "topic": "__eve776561746865722d6167656e74_wkf_workflow_*",
      "consumer": "default",
      "retryAfterSeconds": 5,
      "initialDelaySeconds": 0
    }
  ],
  "environment": {
    "WORKFLOW_PRECONDITION_GUARD": "1"
  }
}

🟠 functions/__server.func • 17 public routes, 1 internal alias • 10.21 MB

Metric	Value
Public routes	`/` `/.well-known/workflow/v1/webhook/[token]` `/eve/v1/activity/[token]` `/eve/v1/callback/[token]` `/eve/v1/connections/[name]/callback/[attemptId]/[token]` `/eve/v1/connections/[name]/callback/[token]` `/eve/v1/health` `/eve/v1/info` `/eve/v1/session` `/eve/v1/session/[parentSessionId]/subagents/[callId]/[childSessionId]/stream` `/eve/v1/session/[sessionId]` `/eve/v1/session/[sessionId]/cancel` `/eve/v1/session/[sessionId]/clear` `/eve/v1/session/[sessionId]/compact` `/eve/v1/session/[sessionId]/reset` `/eve/v1/session/[sessionId]/stream` `/eve/v1/task-input/[token]`
Internal aliases	`/__server`
Runtime	`nodejs24.x`
Handler	`index.mjs`
Payload	10.21 MB
Function files	10.21 MB across 116 files
Traced dependencies	0 B
Signal	🟠 Bundled file `index.mjs` is 2.07 MB (20.3%)

🟠 🔎 Dependency Analysis

📦 Bundled files:

Bundled file size
🟠 index.mjs                        [##########..............] 2.07 MB 20.3%
🟡 _libs/undici.mjs                 [#####...................] 980.8 kB 9.6%
🟡 _chunks/esm-Fqlolk7e.mjs         [###.....................] 723.3 kB 7.1%
🟡 _chunks/chatgpt-model.mjs        [###.....................] 698.7 kB 6.8%
🟡 _chunks/signal-exit-Dsy-TT0V.mjs [###.....................] 616.2 kB 6.0%
🔴 Other bundled files              [########################] 5.11 MB 50.1%

🧾 Vercel Config

{
  "handler": "index.mjs",
  "launcherType": "Nodejs",
  "shouldAddHelpers": false,
  "supportsResponseStreaming": true,
  "runtime": "nodejs24.x"
}

Build Timing: `e2e/fixtures/agent-tools-sandbox`

This is an informational timing measurement inside eve build, from preflight through publication. Output-size measurement and profile writing are excluded.

Build mode: deployable Vercel build with sandbox template prewarm included.

Build pipeline: 4.72 s -> 4.73 s (+7.2 ms) vs main (856d765).
Timing is informational: shared GitHub runners are too variable for a hard timing budget.

Detailed phase timings vs `main (856d765)`

Phase	Baseline	Current	Delta
`extension.check`	0.6 ms	6.9 ms	+6.3 ms
`project.resolve`	0.2 ms	5.7 ms	+5.5 ms
`workspace.create`	0.4 ms	0.9 ms	+0.5 ms
`host.prepare`	417.0 ms	403.6 ms	-13.4 ms
`vercel.service-prefix.resolve`	1.4 ms	1.4 ms	0.0 ms
`nitro.create`	409.5 ms	407.9 ms	-1.6 ms
`sandbox.prewarm`	280.5 ms	340.7 ms	+60.2 ms
`nitro.cache.prepare`	0.2 ms	0.2 ms	0.0 ms
`nitro.prepare`	0.6 ms	0.7 ms	+0.1 ms
`nitro.public-assets`	0.6 ms	0.6 ms	0.0 ms
`nitro.prerender`	0.3 ms	0.3 ms	0.0 ms
`nitro.bundle`	3.54 s	3.50 s	-32.6 ms
`nitro.cache.write`	0.3 ms	0.3 ms	0.0 ms
`vercel.workflow-function.materialize`	56.4 ms	49.5 ms	-6.9 ms
`agent-summary.emit`	0.6 ms	0.5 ms	-0.1 ms
`nitro.close`	0.1 ms	0.1 ms	0.0 ms
`output.publish`	13.3 ms	2.5 ms	-10.8 ms
`workspace.remove`	1.6 ms	1.5 ms	-0.1 ms

e0-gh-vercel-connect · 2026-09-17T07:43:13Z

CI run 35194977969 failed as expected for this red regression PR. The new issue-3458.repro.integration.test.ts has the intended two deterministic failures: cancellation leaves the pending input batch; the subsequent model-call batch runs once instead of twice. The remaining integration suite passed: 871 of 873 tests. This is product behavior under test, not a transient infrastructure failure, so I did not rerun the job.

e0-gh-vercel-connect · 2026-09-17T07:48:13Z

CI run 35195653042 failed only in the openai-sol agent-self-modification evaluation. Three of four evaluations passed, and the failed offer-repair evaluation passed all seven task gates but modified an unrelated fixture file, tests/eval_reorder_plan.test.ts. This is a stochastic model-evaluation failure rather than a failure in the cancellation change. The run was on attempt 1, so I retried the failed jobs.

e0-gh-vercel-connect · 2026-09-17T07:48:56Z

CI run 35195653128 has deterministic test failures, so I did not retry it. The Windows integration suite reports that turn-cancellation.integration.test.ts now emits a turn.cancelled event where the existing session-limit cancellation test expects none. The scenario suite also fails an existing build-extension declaration assertion: it expects DynamicSentinel from eve/tools but received it from eve. The first failure is in the cancellation behavior changed here and needs a code or test expectation decision before CI can pass.

e0-gh-vercel-connect · 2026-09-17T07:50:26Z

CI run 35195653049 failed only in the deployed fixture-tasks evaluation task.dispatch-batch.start.accepted-partial-failure. All 66 behavior gates passed, but the task did not reach completed within the final 30-second verification window. The change does not touch that dispatch path and this is consistent with a timing or infrastructure flake. This was attempt 1, so I retried the failed jobs.

test(eve): reproduce stale input after steered cancellation

eb91ec5

vercel Bot deployed to Preview – eve-pkg September 17, 2026 07:32 View deployment

e0-gh-vercel-connect Bot assigned AndrewBarba Sep 17, 2026

e0-gh-vercel-connect Bot requested a review from AndrewBarba September 17, 2026 07:32

vercel Bot deployed to Preview – eve-docs September 17, 2026 07:33 View deployment

fix(eve): clear pending input on cancelled turns

793c315

e0-gh-vercel-connect Bot marked this pull request as ready for review September 17, 2026 07:39

vercel Bot deployed to Preview – eve-pkg September 17, 2026 07:40 View deployment

vercel Bot deployed to Preview – eve-docs September 17, 2026 07:41 View deployment

e0-gh-vercel-connect Bot mentioned this pull request Sep 17, 2026

Steering a background subagent parked on a proxied approval leaves an unanswerable pending input batch: child parks after one cycle forever, parent never notified #3458

Open

Navigation Menu

test(eve): reproduce stale input after steered cancellation - #3459

test(eve): reproduce stale input after steered cancellation#3459
e0-gh-vercel-connect[bot] wants to merge 2 commits into
mainfrom
e0/issue-3458-pending-input

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

vercel Bot commented Sep 17, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 17, 2026 •

edited

Loading

Package Details

`eve init` install details

Payload Size Graph

Top Function Payloads

Uh oh!

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

e0-gh-vercel-connect Bot commented Sep 17, 2026

Regression test for #3458

Uh oh!

vercel Bot commented Sep 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Sep 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Bundle + Package Summary: apps/fixtures/weather-agent

Delta vs main (856d765)

eve init install

Package Details

eve init install details

Payload Size Graph

Top Function Payloads

Build Timing: e2e/fixtures/agent-tools-sandbox

Uh oh!

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

e0-gh-vercel-connect Bot commented Sep 17, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

vercel Bot commented Sep 17, 2026 •

edited

Loading

github-actions Bot commented Sep 17, 2026 •

edited

Loading

Bundle + Package Summary: `apps/fixtures/weather-agent`

Delta vs `main (856d765)`

`eve init` install

`eve init` install details

Build Timing: `e2e/fixtures/agent-tools-sandbox`