• fix(test): stabilize lifecycle, sleep, queue, and run edge cases #4660 : 2 dependent PRs (#4664 , #4665 )
• test(rivetkit): re-enable gateway URL and direct-registry coverage #4659
• fix(rivetkit): add lifecycle error retry and gateway HTTP routing #4666
• fix(rivetkit-core): remove legacy timeouts #4799
• chore: add CLAUDE.md guidance on test fixtures, SDK regen, and TS client layer #4796
• fix(inspector): stream patched state updates #4795
• fix(rivetkit): refresh stale actor handles #4794
• test(rivetkit-napi): move private tests to crate root #4793
• test(rivetkit-core): move private tests to crate root #4792 👈 (View in Graphite)
• test(runner-configs): assert envoy protocol version #4791
• fix(serverless): split metadata protocol errors #4789 : 1 other dependent PR (#4790 )
• chore(sqlite): add open close lifecycle for envoy actors #4788
• fix(rivetkit): share lifecycle shutdown grace #4810
• fix(pegboard): move sqlite v1 migration into actor workflow #4787 : 1 other dependent PR (#4797 )
• test(rivetkit): cover c.db access from onWake/onSleep/onDestroy #4786
• fix(rivetkit-napi): run onMigrate before createState/onCreate/createVars #4785
• refactor(rivetkit-core): merge ready+started into lifecycle_started, drop redundant napi calls #4784
• fix(rivetkit): drain native sleep side tasks reliably #4782 : 1 other dependent PR (#4783 )
• fix(rivetkit): route raw request fetches to actors #4781
• fix(pegboard): refresh runner config after envoy connect #4778 : 2 other dependent PRs (#4779 , #4798 )
• chore(kitchen-sink): configurable endpoint #4766 : 2 other dependent PRs (#4754 , #4777 )
• refactor(rivetkit-core): rename StopReason to ShutdownKind and kill enum fallthroughs #4765
• chore(envoy-client): log command and event payloads #4751
• fix(rivetkit): minor sleep-cleanup follow-ups #4761
• docs(rivetkit): clean up stale setPreventSleep references in docs and JSDoc #4760
• docs(rivetkit): document keepAwake/waitUntil counter-arm race and surface bridge errors #4759
• test(engine): force overdue branch in alarm-during-sleep test with negative offset #4758
• fix(rivetkit-core): return stopping not starting when sleep/destroy called mid-shutdown #4757
• test(rivetkit): remove sleepWithPreventSleep fixture and tests #4756
• test(engine): add regression test for alarm-during-sleep-transition race #4755
• test(rivetkit-core): cover sleep/destroy guards and prevent_sleep no-op #4750
• docs: add keepAwake vs waitUntil table and internal sleep-sequence invariants #4749
• feat(rivetkit): expose c.keepAwake(promise) through native path #4748
• refactor(rivetkit): deprecate setPreventSleep to no-op in TypeScript #4747
• fix(rivetkit-core): enumerate non-drained counters in grace deadline warn log #4746
• fix(rivetkit-core): abort graceful cleanup task when grace deadline elapses #4745
• fix(rivetkit-core): error on sleep/destroy before startup or already-requested #4743
• fix(rivetkit-core): cap SerializeState shutdown timeout at 15s #4742
• refactor(rivetkit-core): remove prevent_sleep field and deprecate set_prevent_sleep to no-op #4741
• fix(api-public): remove orphan import_export actors module #4764
• main

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

PR #4792 Review: Move private tests to crate root

Overview

Migrates test modules from inline #[cfg(test)] mod tests { ... } blocks in source files (or tests/modules/) to flat files under tests/ at the crate root. Because these tests need access to private items, autotests = false prevents Cargo from compiling the tests/*.rs files as standalone integration test binaries; they remain included via #[path] shims in each source file. The PR also centralizes test_inspector_env_lock() into src/inspector/auth.rs and fixes a test assertion in task.rs.

This directly implements the CLAUDE.md requirement: "Rust tests live under tests/, not inline #[cfg(test)] mod tests in src/."

What's Good

• CLAUDE.md compliance: All inline test blocks are now in tests/, matching the codebase convention.
• autotests = false is correct: Without it, Cargo would try to compile the #[path]-included files as standalone integration test binaries and fail on use super::*;. The setting is a necessary consequence of the design.
• Centralized test_inspector_env_lock(): Previously a static INSPECTOR_ENV_LOCK local to tests/modules/inspector.rs, it is now a shared OnceLock<Mutex<()>> in src/inspector/auth.rs. After flattening, inspector auth tests can land in different test binaries; the shared lock ensures proper env-var serialization across all tests touching _RIVET_TEST_INSPECTOR_TOKEN.
• Test assertion fix in task.rs: Correcting the assertion from 0 to 1 test_batch_get_call_count() (with a clarifying comment) and renaming the test to _without_startup_bundle_batch_get improves accuracy.
• Consistent shim pattern: Every source file uses the same #[path = "../../tests/X.rs"] shim with an explanatory comment.

Issues

1. autotests = false has no explanation in Cargo.toml (minor)

The setting is non-obvious. A future contributor adding a standalone integration test would be surprised it is silently ignored. A comment prevents this:

# Tests in tests/ are included via #[path] shims in source files for private-item access.
# They are not standalone integration test binaries, so autotests must stay false.
autotests = false

2. test_inspector_env_lock() uses std::sync::Mutex instead of parking_lot::Mutex (minor)

CLAUDE.md says test-only global serialization should use parking_lot. Even though the function is #[cfg(test)], it lives in a production source file, and the codebase consistently uses parking_lot for forced-sync contexts:

#[cfg(test)] use parking_lot::Mutex;
#[cfg(test)] use std::sync::OnceLock;

#[cfg(test)]
pub(crate) fn test_inspector_env_lock() -> &'static Mutex<()> {
    static LOCK: OnceLock<Mutex<()>> = OnceLock::new();
    LOCK.get_or_init(|| Mutex::new(()))
}

3. mod moved_tests is inconsistently applied and transitionally named (minor)

Some files wrap tests in mod moved_tests { ... } (inspector.rs, registry_http.rs, serverless.rs, task.rs, work_registry.rs) while others do not. The name describes migration history rather than module purpose. Post-migration these wrappers add no value; either remove them or rename to something semantically meaningful.

4. Substantive test fix bundled into a mechanical refactor (low risk, worth noting)

The task.rs change corrects the expected batch-get count from 0 to 1 and adds an env guard. This is a genuine correctness fix (the old assertion was wrong; the test was previously flaky when _RIVET_TEST_INSPECTOR_TOKEN was set by a preceding test). Good change, but non-trivial inside an otherwise purely structural refactor.

Summary

The approach is sound and the CLAUDE.md requirement is satisfied. The two actionable items are the missing Cargo.toml comment and the std::sync::Mutex to parking_lot::Mutex swap in the test helper. The rest are nits.

Preview packages published to npm

Install with:

npm install rivetkit@pr-4792

All packages published as 0.0.0-pr.4792.8c20f1a with tag pr-4792.

Engine binary is shipped via @rivetkit/engine-cli on linux-x64-musl, linux-arm64-musl, darwin-x64, and darwin-arm64. Windows users should use the release installer or set RIVET_ENGINE_BINARY.

Docker images:

docker pull rivetdev/engine:slim-8c20f1a
docker pull rivetdev/engine:full-8c20f1a

npm install rivetkit@pr-4792
npm install @rivetkit/react@pr-4792
npm install @rivetkit/rivetkit-napi@pr-4792
npm install @rivetkit/workflow-engine@pr-4792