[Gladorme]

Description

Upgrade Rsbuild to v2 and enable the React Compiler through pluginReact({ reactCompiler }) for the module federation bundles and through .swcrc for the library build. Plugins target React 18 and depend on react-compiler-runtime.

Screenshots

Checklist

• Pull request has a descriptive title and context useful to a reviewer.
• Pull request title follows the [<catalog_entry>] <commit message> naming convention using one of the
  following catalog_entry values: FEATURE, ENHANCEMENT, BUGFIX, BREAKINGCHANGE, DOC,IGNORE.
• All commits have DCO signoffs.

UI Changes

• Changes that impact the UI include screenshots and/or screencasts of the relevant changes.
• Code follows the UI guidelines.

[Nexucis]

🤖 This review was generated by GitHub Copilot (AI assistant), on behalf of the reviewer.

Thanks for this, overall it looks good to me 👍 A few non-blocking suggestions inline.

What was verified (PR checked out locally, compared against origin/main)

• All 33 plugin packages get react-compiler-runtime + pluginReact({ reactCompiler }) (only e2e is untouched, as expected).
• The compiler is effectively applied in both outputs:
  • ESM lib (build:esm via .swcrc): in table, 15/30 emitted files import { c as _c } from "react-compiler-runtime" (the only non-compiled file with JSX is bootstrap.js, which is fine).
  • MF bundles (rsbuild build): memo_cache_sentinel present in the output.
• react-compiler-runtime is correctly in dependencies (the published lib/ imports it), and its peer range (react ^17 || ^18 || ^19) matches.
• No need to add it to MF shared: it's stateless and its react import resolves to the shared singleton.
• tsc -p tsconfig.json passes on rsbuild.shared.ts + all rsbuild.config.ts against Rsbuild v2 types; no removed v2 options in use.
• Bundle size (table MF build): 8328 → 8343 kB raw / 3478 → 3467 kB gzip, i.e. negligible.
• The oxlint config already enforces the compiler-derived rules (react/hooks, react/memo-dependencies, react/unsupported-syntax, react/todo…) and there's no use no memo / hooks-rule suppression in the sources, so the codebase is in good shape for this.

Other notes

• Vitest does not run the compiler, so unit tests exercise the uncompiled code; compiler-specific regressions would only be caught by e2e/manual testing. Lint rules mitigate this, but it may be worth mentioning in the release notes.
• Minor: the PR checklist is unchecked and the screenshot has no caption explaining what it shows.