iframe-proxy

danielroe · 2026-05-07T20:24:48Z

🔗 Linked issue

resolves #32102
resolves #32437

📚 Description

with watch: false, useFetch was unwrapping the key to a plain string before passing it to useAsyncData, which froze the 'key'

that's what's behind both linked issues: getCachedData always saw the original key, and two useFetch calls that started out with the same shape kept sharing one async data slot even after their inputs diverged - so execute() on one overwrote the other's data.

this came from #32019, which fixed watch: false not preventing auto-refetch (#32016) but disabled key reactivity entirely.

so this pr passes the reactive key through, and adds an internal _keyTriggersExecute flag on useAsyncData's key watcher. when false, the data slot is still migrated to the new key (so getCachedData and manual execute() see it), but the auto-execute on key change is skipped. useFetch sets it to false whenever watch: false.

reactive query/params still don't trigger auto-refetch.

bolt-new-by-stackblitz · 2026-05-07T20:24:51Z

coderabbitai

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@test/nuxt/use-fetch.test.ts`:
- Around line 372-390: Two effectScope instances (scope1 and scope2 created via
effectScope() and run(createScope())) are not stopped, risking cross-test
leakage; fix by calling scope1.stop() and scope2.stop() after the test actions
(after the awaits and assertions) to dispose watchers/deps created by
createScope(); ensure you add the stop calls referencing the exact variables
scope1 and scope2 so both scopes are cleaned up regardless of assertion outcomes
(or wrap in finally if you prefer deterministic teardown).

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

Push a commit to this branch (recommended)
Create a new PR with the fixes

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 00b8dc6c-6e56-4ca9-8ea2-10b9bc232a37

📥 Commits

Reviewing files that changed from the base of the PR and between 8630e8f and d97875b.

📒 Files selected for processing (3)

packages/nuxt/src/app/composables/asyncData.ts
packages/nuxt/src/app/composables/fetch.ts
test/nuxt/use-fetch.test.ts

coderabbitai · 2026-05-07T20:27:23Z

+    const scope1 = effectScope()
+    const { data: d1, id: i1, query: q1, execute: e1 } = scope1.run(createScope())!
+    const scope2 = effectScope()
+    const { data: d2 } = scope2.run(createScope())!
+
+    expect.soft(d1.value).toStrictEqual(undefined)
+    expect.soft(d2.value).toStrictEqual(undefined)
+
+    q1.value.someInput = 'test'
+    i1.value++
+    await e1()
+    await flushPromises()
+    await nextTick()
+    await flushPromises()
+
+    // d1 should have data; d2 should remain untouched
+    expect.soft(d1.value).toMatchObject({ url: '/api/key/2' })
+    expect.soft(d2.value).toStrictEqual(undefined)
+  })


⚠️ Potential issue | 🟡 Minor | ⚡ Quick win

Stop created effect scopes to avoid cross-test leakage.

At Line 372 and Line 374, two effectScope() instances are created but never disposed. This can leak watchers/deps into subsequent tests and make outcomes order-dependent.

Proposed fix

const scope1 = effectScope() const { data: d1, id: i1, query: q1, execute: e1 } = scope1.run(createScope())! const scope2 = effectScope() const { data: d2 } = scope2.run(createScope())! - expect.soft(d1.value).toStrictEqual(undefined) - expect.soft(d2.value).toStrictEqual(undefined) + try { + expect.soft(d1.value).toStrictEqual(undefined) + expect.soft(d2.value).toStrictEqual(undefined) - q1.value.someInput = 'test' - i1.value++ - await e1() - await flushPromises() - await nextTick() - await flushPromises() + q1.value.someInput = 'test' + i1.value++ + await e1() + await flushPromises() + await nextTick() + await flushPromises() - // d1 should have data; d2 should remain untouched - expect.soft(d1.value).toMatchObject({ url: '/api/key/2' }) - expect.soft(d2.value).toStrictEqual(undefined) + // d1 should have data; d2 should remain untouched + expect.soft(d1.value).toMatchObject({ url: '/api/key/2' }) + expect.soft(d2.value).toStrictEqual(undefined) + } finally { + scope1.stop() + scope2.stop() + }

📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change

const scope1 = effectScope()

const { data: d1, id: i1, query: q1, execute: e1 } = scope1.run(createScope())!

const scope2 = effectScope()

const { data: d2 } = scope2.run(createScope())!

expect.soft(d1.value).toStrictEqual(undefined)

expect.soft(d2.value).toStrictEqual(undefined)

q1.value.someInput = 'test'

i1.value++

await e1()

await flushPromises()

await nextTick()

await flushPromises()

// d1 should have data; d2 should remain untouched

expect.soft(d1.value).toMatchObject({ url: '/api/key/2' })

expect.soft(d2.value).toStrictEqual(undefined)

})

const scope1 = effectScope()

const { data: d1, id: i1, query: q1, execute: e1 } = scope1.run(createScope())!

const scope2 = effectScope()

const { data: d2 } = scope2.run(createScope())!

try {

expect.soft(d1.value).toStrictEqual(undefined)

expect.soft(d2.value).toStrictEqual(undefined)

q1.value.someInput = 'test'

i1.value++

await e1()

await flushPromises()

await nextTick()

await flushPromises()

// d1 should have data; d2 should remain untouched

expect.soft(d1.value).toMatchObject({ url: '/api/key/2' })

expect.soft(d2.value).toStrictEqual(undefined)

} finally {

scope1.stop()

scope2.stop()

}

})

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@test/nuxt/use-fetch.test.ts` around lines 372 - 390, Two effectScope instances (scope1 and scope2 created via effectScope() and run(createScope())) are not stopped, risking cross-test leakage; fix by calling scope1.stop() and scope2.stop() after the test actions (after the awaits and assertions) to dispose watchers/deps created by createScope(); ensure you add the stop calls referencing the exact variables scope1 and scope2 so both scopes are cleaned up regardless of assertion outcomes (or wrap in finally if you prefer deterministic teardown).

pkg-pr-new · 2026-05-07T20:27:33Z

Open in StackBlitz

@nuxt/kit

npm i https://pkg.pr.new/@nuxt/kit@35002

@nuxt/nitro-server

npm i https://pkg.pr.new/@nuxt/nitro-server@35002

nuxt

npm i https://pkg.pr.new/nuxt@35002

@nuxt/rspack-builder

npm i https://pkg.pr.new/@nuxt/rspack-builder@35002

@nuxt/schema

npm i https://pkg.pr.new/@nuxt/schema@35002

@nuxt/vite-builder

npm i https://pkg.pr.new/@nuxt/vite-builder@35002

@nuxt/webpack-builder

npm i https://pkg.pr.new/@nuxt/webpack-builder@35002

commit: 31ab871

codspeed · 2026-05-07T20:34:11Z

Merging this PR will not alter performance

✅ 20 untouched benchmarks
⏩ 3 skipped benchmarks¹

_{Comparing fix/usefetch-stale-key (31ab871) with main (778f190)}

3 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

OrbisK · 2026-05-08T10:17:48Z


+      if (watchSources === false) {
+        // opt-out of automatic re-execution while keeping key reactive
+        ;(_asyncDataOptions as typeof _asyncDataOptions & { _keyTriggersExecute?: boolean })._keyTriggersExecute = false


Why cant this be an "official" option for async data (AsyncDataOptions)?

you mean, watch: false?

if you think it's a good idea, let's open a new PR to add the feature, but we can still get this one in as a fix for the next patch

danielroe added 2 commits May 7, 2026 10:17

test: add failing useFetch tests

c0481a7

fix(nuxt): update useFetch key even with watch: false

d97875b

danielroe requested a review from OrbisK May 7, 2026 20:24

github-actions Bot added 5.x 🐛 bug Something isn't working as expected labels May 7, 2026

This comment has been minimized.

Sign in to view

coderabbitai Bot reviewed May 7, 2026

View reviewed changes

Merge branch 'main' into fix/usefetch-stale-key

31ab871

danielroe self-assigned this May 7, 2026

OrbisK reviewed May 8, 2026

View reviewed changes

OrbisK approved these changes May 11, 2026

View reviewed changes

danielroe merged commit 68ebb99 into main May 11, 2026
31 checks passed

danielroe deleted the fix/usefetch-stale-key branch May 11, 2026 11:25

This was referenced May 11, 2026

v5.0.0 #34497

Draft

Need to port from upstream nuxt/bridge#1813

Open

danielroe added a commit that referenced this pull request May 11, 2026

fix(nuxt): update useFetch key even with watch: false (#35002)

83e7351

github-actions Bot mentioned this pull request May 11, 2026

v4.4.6 #35030

Merged

davidstackio pushed a commit to davidstackio/nuxt that referenced this pull request May 30, 2026

fix(nuxt): update useFetch key even with watch: false (nuxt#35002)

d14a634

Sunbelt Computer Software

PL/B Language Development and Support

Navigation Menu

Repository navigation

fix(nuxt): update `useFetch` key even with `watch: false` - #35002

fix(nuxt): update `useFetch` key even with `watch: false`#35002
danielroe merged 3 commits into
mainfrom
fix/usefetch-stale-key

danielroe commented May 7, 2026

Uh oh!

bolt-new-by-stackblitz Bot commented May 7, 2026

Uh oh!

This comment has been minimized.

coderabbitai Bot left a comment

Uh oh!

coderabbitai Bot May 7, 2026

Uh oh!

pkg-pr-new Bot commented May 7, 2026 •

edited

Loading

Uh oh!

codspeed Bot commented May 7, 2026 •

edited

Loading

Uh oh!

OrbisK May 8, 2026 •

edited

Loading

Uh oh!

danielroe May 9, 2026

Uh oh!

danielroe May 10, 2026

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Sunbelt Computer Software

PL/B Language Development and Support

Conversation

danielroe commented May 7, 2026

🔗 Linked issue

📚 Description

Uh oh!

bolt-new-by-stackblitz Bot commented May 7, 2026

Uh oh!

This comment has been minimized.

coderabbitai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

coderabbitai Bot May 7, 2026

Choose a reason for hiding this comment

Uh oh!

pkg-pr-new Bot commented May 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codspeed Bot commented May 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will not alter performance

Footnotes

Uh oh!

OrbisK May 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

danielroe May 9, 2026

Choose a reason for hiding this comment

Uh oh!

danielroe May 10, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

pkg-pr-new Bot commented May 7, 2026 •

edited

Loading

codspeed Bot commented May 7, 2026 •

edited

Loading

OrbisK May 8, 2026 •

edited

Loading