[astro-factory]

Changes

• Failed fetch() calls in the generated server island client script no longer surface as Uncaught (in promise) TypeError. Previously, when a fetch rejected (network drop, aborted navigation, blocked cross-origin redirect), the top-level await in the island module would throw unhandled, polluting error trackers even though the fallback content remained visible. Both the GET and POST island fetch calls now append .catch(() => {}), and replaceServerIsland returns early on a missing response or a failed r.text() read, keeping the fallback silently.
• The return statement suggested in the issue cannot be used at module top-level; the .catch(() => {}) approach achieves the same effect without a syntax error.

Testing

• Extended packages/astro/test/units/server-islands/server-islands-render.test.ts to cover the case where fetch rejects, asserting that no error is thrown and the fallback content is preserved.

Docs

• No docs update needed; this is an internal error-handling fix with no new public API or configuration surface.

Closes #18495

[changeset-bot]

🦋 Changeset detected

Latest commit: 2012a86

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

This PR includes changesets to release 1 package

Name	Type
astro	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

[astro-factory]

astro-author owns this pull request

Factory's code author persona works on this branch when a maintainer or Factory's reviewer requests changes, or when checks fail: it pushes follow-up commits, replies to review threads, and asks for another review. It never merges. Unassign astro-author to take the pull request back; reassign it to start a fresh budget.

Rounds used: 0 of 5
Status: waiting for feedback.

_{— astro-author (Factory persona)}

[astro-factory]

Verdict: approved. No blocking findings remain.

_{GitHub doesn't let Factory approve a pull request it opened, so this verdict is recorded as a comment.}

Findings

No findings.

Merge Requirement Assessment

Status: satisfied

• A rejected island fetch() (GET or POST) must not cause an uncaught TypeError and must keep the fallback (#18495): Both generated fetch calls in ServerIslandComponent.getIslandContent() now end with .catch(() => {}), so response becomes undefined when the request fails. replaceServerIsland gains a !r guard, so it returns before touching the DOM. The new parameterized unit tests run the generated GET and POST scripts with a rejecting fetch. Without the fix, the top-level await would reject runIslandScript, so these tests would catch a regression.
• A failed r.text() read must keep the fallback: The body read is now wrapped in try { ... } catch { return; } before any DOM change. This matters because replaceServerIsland(...) isn't awaited in the island script, so a rejection there would otherwise surface as an unhandled promise rejection. The new renderServerIslandRuntime test covers it with a Response whose text() rejects. The runtime's basic minification (trim lines, drop // lines, join with spaces) leaves the new multi-line try/catch valid, and the optional catch binding is ES2019 syntax, which the browsers that already run this module script support.

Review Status

Verdict: ready to merge based on static review
Context: PR #18497 plus linked issue #18495 (as described in the PR body)
Scope: PR base 5fe699c to head 2012a86 (3 files).
Changeset: present and covers astro (patch).
Limitations: The linked issue's body and comments were not retrieved separately; the requirements come from the PR description.
Validation: Static review only. No project code, tests, builds or checks were run.
GitHub context: collected
Git scope: PR base/head verified

This review was made by an LLM. The analysis may be wrong, and reports might be incorrect.

[codspeed]

Merging this PR will not alter performance

1 warning, 1 note

⚠️ 3 benchmarks are dominated by system calls

System calls cannot be simulated consistently and are left out of the measure. The reported value understates the real cost. Switch to the Walltime instrument to measure them.

Learn more about system calls

See affected benchmarks

ℹ️ 1 benchmark has no flamegraph

The flamegraph wasn't generated: most of the time was spent in runtime internals, which would leave nothing useful to analyze.

See affected benchmarks

✅ 21 untouched benchmarks

---

_{Comparing factory/fix-18495 (2012a86) with main (d145d19)¹}

Footnotes

1. No successful run was found on main (5fe699c) during the generation of this report, so d145d19 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩