revert(rivetkit-core): drop dangling http body stream re-exports

[abcxff]

No description provided.

[abcxff]

Stack for rivet-dev/rivet

• #5530 fix(rivetkit-core): cancel driver alarm before sqlite teardo …
• #5531 feat(container-runner): keep instance warm instead of self-e …
• #5533 chore(container-runner): default stop grace to 10s
• #5528 fix(rivetkit): log and surface actor startup failures
• #5529 fix(rivetkit): treat empty actor input bytes as absent
• #5532 revert(rivetkit-core): drop dangling http body stream re-exp … 👈
• main

Get stack: forklift get 5532
Push local edits: forklift submit
Merge when ready: forklift merge 5532

_{change nwmvwykt}

[railway-app]

🚅 Deployed to the rivet-pr-5532 environment in rivet-frontend

Service	Status	Web	Updated (UTC)
kitchen-sink	✅ Success (View Logs)	Web	Jul 31, 2026 at 4:14 am
frontend-inspector	✅ Success (View Logs)	Web	Jul 31, 2026 at 4:13 am
frontend-cloud	✅ Success (View Logs)	Web	Jul 31, 2026 at 4:13 am
website	✅ Success (View Logs)	Web	Jul 31, 2026 at 4:13 am
mcp-hub	✅ Success (View Logs)	Web	Jul 31, 2026 at 4:13 am
ladle	✅ Success (View Logs)	Web	Jul 31, 2026 at 4:13 am

[claude]

Review

Small, correct cleanup. I verified the underlying claim: engine/sdks/rust/envoy-client/src/config.rs no longer defines HttpRequestBodyStream or HTTP_BODY_STREAM_CHANNEL_CAPACITY — only HttpRequest, HttpResponse, ResponseChunk, EnvoyConfig, etc. remain. So the removed pub use items in rivetkit-rust/packages/rivetkit-core/src/lib.rs were re-exporting nonexistent paths, which would fail to compile as-is. Dropping them and keeping ResponseChunk (which still exists in config.rs) is the right fix.

A few notes, none blocking this PR:

• Related dangling reference elsewhere: rivetkit-typescript/packages/rivetkit-napi/tests/actor_factory.rs still imports rivetkit_core::HttpRequestBodyStream as CoreHttpRequestBodyStream and use super::{..., HttpRequestBodyStream, ...}, and constructs CoreHttpRequestBodyStream::new(...). Neither rivetkit_core::HttpRequestBodyStream nor a module-local HttpRequestBodyStream exist anywhere in the current tree (checked actor_factory.rs src, which doesn't define or import that name). That test file looks like it was already broken before this change — this PR doesn't cause it, but since it's the same identifier being cleaned up here, it'd be worth a follow-up to either restore the type where the tests expect it or delete/update the now-stale test code.
• Test coverage: this is a pure re-export removal with no behavior change, so no new tests are needed — makes sense that none were added.
• PR description is empty. Per repo convention, even a one-line "why" bullet is nice to have, though for a mechanical revert like this it's low value.

No security, performance, or logging concerns — this is just narrowing a public re-export surface to match what actually exists upstream.