[harsh21234i]

Fixes #7426, where calling a local client-state setter from a sibling component could raise a ReferenceError and leave the displayed value unchanged.

The compiler now detects when local client state is shared across components and keeps those components in a compatible React scope. Added a compiler regression test and a Playwright test that verifies the sibling setter updates the rendered value. Included changelog entries for both affected packages.

Testing: Targeted unit and Playwright tests, Ruff, Pyright, and git diff --check pass locally.

[cubic-dev-ai]

All reported issues were addressed across 7 files

<sub>Reply with feedback, questions, or to request a fix.

Re-trigger cubic</sub>

[greptile-apps]

Confidence Score: 4/5

[Medium impact] The change affects page compilation for pages that use local client state.

The PR needs one more fix before merging: default styles can still split shared local state.

Findings

1.  Default styles stop updating ▶

Summary

Keeps sibling components that share local client state in the same React scope.

• Local client-state readers and setters stay in one React scope.

_{Reviews (16) · Last reviewed commit: "fix: preserve empty component style scop..." · Reviewed by Greptile}

[codspeed]

Merging this PR will degrade performance by 5.5%

2 warnings

⚠️ 14 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

⚠️ Base and head ran on different runner types

Some benchmarks with significant performance changes were compared across different runner types, so the changes might not come from the code. Run base and head on the same runner type, or use CodSpeed Macro Runners for more stable results.

Learn more about Macro Runners

Open the report in CodSpeed to investigate

❌ 8 regressed benchmarks
✅ 140 untouched benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
❌	test_compile_all_artifacts[_complicated_page]	130.3 ms	142.2 ms	-8.4%
❌	test_compile_all_artifacts[_stateful_page]	38 ms	40.3 ms	-5.7%
❌	test_compile_page[_complicated_page]	194.1 ms	205.8 ms	-5.69%
❌	test_compile_page_full_context[_complicated_page]	210.7 ms	222.6 ms	-5.34%
❌	test_compile_page_full_context[_stateful_page]	45.9 ms	48.4 ms	-5.08%
❌	test_compile_page[_stateful_page]	43.4 ms	45.6 ms	-4.99%
❌	test_compile_all_artifacts[_repeated_stateful_page]	446.5 ms	468.1 ms	-4.62%
❌	test_compile_page[_repeated_stateful_page]	519.2 ms	541.3 ms	-4.1%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing harsh21234i:fix/7426-local-client-state-sibling (83062a4) with main (a9b6613)}

[cubic-dev-ai]

All reported issues were addressed across 4 files (changes from recent commits).

<sub>Reply with feedback, questions, or to request a fix.

View guided diff | Re-trigger cubic</sub>

[cubic-dev-ai]

All reported issues were addressed across 1 file (changes from recent commits).

<sub>Tip: Review your code locally with the cubic CLI to iterate faster.

View guided diff | Re-trigger cubic</sub>

[masenf]

i'm not too keen on patching memo plugin so deeply to support the old experimental client_state.

the new client state already uses scoped access on the javascript side and just works transparently without all this compiler machinery.

[cubic-dev-ai]

All reported issues were addressed across 4 files (changes from recent commits).

<sub>Reply with feedback, questions, or to request a fix.

View guided diff | Re-trigger cubic</sub>

[harsh21234i]

Thanks, I understand the concern. I’ve simplified the prepass based on the review feedback, but this still adds compiler logic for the legacy experimental API. Since the newer client-state API already handles scoped access, would you prefer we keep a compatibility fix for this existing behavior, or close/defer this issue and direct users to the newer API?

[greptile-apps]

Default styles stop updating

Local state added by add_style() is missing from this scan. For example, subclass AutoScroll, return {"color": local.value} from add_style(), and place it beside a local.value reader and setter. The root scan runs before the child's default style is applied and only accounts for App.style. The child can then get its own memo component and state, so clicking the outside setter leaves its color unchanged.

Include default-style state users before computing the shared-scope decisions.