fix(rivetkit): prevent Turbopack from tracing sqlite-vfs dynamic import

[NathanFlurry]

Description

Please include a summary of the changes and the related issue. Please also include relevant motivation and context.

Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Please describe the tests that you ran to verify your changes.

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

[NathanFlurry]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• fix(rivetkit): remove drizzle-orm peer dep and add runtime version check #4492
• chore(skill): add version check to rivetkit skill #4491
• fix(rivetkit): prevent Turbopack from tracing sqlite-vfs dynamic import #4490 👈 (View in Graphite)
• fix(rivetkit): warn when manager port falls back to a different port #4489
• fix(rivetkit): clear connectableTs on actor sleep #4488
• fix(rivetkit): gate node:sqlite behind RIVETKIT_NODE_SQLITE env var #4487
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[railway-app]

🚅 Deployed to the rivet-pr-4490 environment in rivet-frontend

Service	Status	Web	Updated (UTC)
website	😴 Sleeping (View Logs)	Web	Mar 24, 2026 at 2:17 am
frontend-cloud	😴 Sleeping (View Logs)	Web	Mar 24, 2026 at 2:09 am
frontend-inspector	😴 Sleeping (View Logs)	Web	Mar 24, 2026 at 2:06 am
ladle	❌ Build Failed (View Logs)	Web	Mar 24, 2026 at 2:01 am
kitchen-sink	❌ Build Failed (View Logs)	Web	Mar 24, 2026 at 2:00 am
mcp-hub	✅ Success (View Logs)	Web	Mar 24, 2026 at 1:58 am

[claude]

PR Review: fix(rivetkit): prevent Turbopack from tracing sqlite-vfs dynamic import

This is a targeted fix for a Turbopack-specific bundler issue. The change is small and well-reasoned.

What the fix does

By constructing the module specifier as ["@rivetkit", "sqlite-vfs"].join("/") instead of a static string literal, the dynamic import becomes opaque to Turbopack's static analysis at compile time. This prevents Turbopack from eagerly tracing into @rivetkit/sqlite-vfs's transitive WASM dependencies, which it can't handle.

Positives

• The comment clearly explains the "why" and the mechanism — important for a non-obvious workaround like this.
• The fix is minimal and surgical — exactly the right scope.

Minor concern: inline type annotation length

The type annotation on line 31 is quite long and wraps awkwardly:

({ SqliteVfsPool }: { SqliteVfsPool: new (opts: { actorsPerInstance: number; idleDestroyMs: number }) => { acquire(actorId: string): Promise<ISqliteVfs>; shutdown(): Promise<void> } }) =>

Since the file already imports ISqliteVfs from @rivetkit/sqlite-vfs, consider extracting a local type alias for readability:

type SqliteVfsPoolCtor = new (opts: { actorsPerInstance: number; idleDestroyMs: number }) => {
    acquire(actorId: string): Promise<ISqliteVfs>;
    shutdown(): Promise<void>;
};

this.#poolPromise = import(specifier).then(
    ({ SqliteVfsPool }: { SqliteVfsPool: SqliteVfsPoolCtor }) =>
        new SqliteVfsPool({ actorsPerInstance: poolConfig.actorsPerInstance, idleDestroyMs: poolConfig.idleDestroyMs }),
);

This is a readability nit only — the current code is functionally correct.

Questions / considerations

• Fragility of the workaround: Is there a Turbopack issue filed upstream? The Array.join() trick is clever but could theoretically be optimized through in a future Turbopack version. A comment linking to an upstream issue would help future maintainers know if/when this can be removed.
• Other bundlers: This should be safe for webpack/esbuild since they generally don't statically analyze dynamic imports with computed specifiers, but worth verifying.

Summary

The fix is correct and well-commented. The only suggestion is a minor readability improvement on the inline type annotation. LGTM.