feat: add sleep and reschedule endpoints for actors

[MasterPtato]

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

[railway-app]

🚅 Deployed to the rivet-pr-4556 environment in rivet-frontend

Service	Status	Web	Updated (UTC)
kitchen-sink	❌ Build Failed (View Logs)	Web	Apr 5, 2026 at 11:41 am
mcp-hub	✅ Success (View Logs)	Web	Apr 4, 2026 at 11:05 pm
frontend-inspector	❌ Build Failed (View Logs)	Web	Apr 4, 2026 at 11:05 pm
frontend-cloud	❌ Build Failed (View Logs)	Web	Apr 4, 2026 at 11:04 pm
website	❌ Build Failed (View Logs)	Web	Apr 4, 2026 at 11:04 pm
ladle	❌ Build Failed (View Logs)	Web	Apr 4, 2026 at 11:04 pm

[MasterPtato]

• chore: remove global epoxy contention #4554
• feat: native SQLite backend via napi-rs with KV channel protocol #4496
• feat: actor query in gateway #4553
• feat: add sleep and reschedule endpoints for actors #4556 👈 (View in Graphite)
• fix: optimize e2e actor path #4551
• feat(pb): actor v1 to v2 migration #4548
• feat(envoy-client): fully flesh out tunnel impl #4547
• feat: basic rivetkit impl #4544
• feat: basic envoy tunnel impl #4534
• chore: envoy client #4521 : 1 other dependent PR (#4531 )
• feat(pb): actors v3 #4463
• feat(ups): implement queue subs #4486
• feat(cache): add in flight deduping #4459
• fix(cache): clean up lib #4457
• fix(config): allow configuring topo dcs via map, fix pg ssl mode config #4456
• fix(api): dont 500 on cross dc response #4449
• chore: update byte units on frontend #4444
• fix: runner alloc idx logic, api auth for actor get #4443
• main

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

[claude]

PR Review: feat: add sleep and reschedule endpoints for actors

Summary

This PR adds two new HTTP API endpoints (POST /actors/{actor_id}/sleep and POST /actors/{actor_id}/reschedule), renames the internal Reallocate signal to Reschedule throughout pegboard, and introduces a new "stopping" SSE event to allow the serverless actor stream to close early when an actor stops gracefully.

The feature design is architecturally sound and follows existing patterns well, but there are a few issues that should be addressed before merging.

---

Issues

High: Signal rename violates protocol versioning rules

engine/packages/pegboard/src/workflows/actor2/mod.rs renames the signal name from "pegboard_actor2_reallocate" to "pegboard_actor2_reschedule". Per CLAUDE.md, published signal/protocol names must not be modified in-place — a new versioned schema must be added and the old one bridged forward. Any actors currently in a GoingAway state waiting on the old signal name will never receive the renamed signal, which is a correctness risk for rolling deployments.

Additionally, there's now a naming inconsistency: runtime.rs still uses try_reallocate, going_away, and comments like // Always immediately reallocate if going away, while the signal itself is now Reschedule. The Transition::Reallocating enum variant is also unchanged.

Medium: Missing await on stream.writeSSE for the "stopping" event

In rivetkit-typescript/packages/rivetkit/src/drivers/engine/actor-driver.ts (around line 506) and engine/sdks/typescript/test-envoy/src/index.ts (around line 201), the "stopping" SSE event write is missing await:

// Missing await — should be: await stream.writeSSE(...)
stream.writeSSE({ event: "stopping", data: "" });

Hono's SSEStreamingApi.writeSSE returns Promise<void>. Without awaiting, the event may not be flushed before the stream closes, making the feature silently non-functional in some cases. The ping loop in the same file correctly uses await stream.writeSSE(...).

Medium: No tests

The new sleep/reschedule endpoints and the "stopping" SSE flow have no test coverage. The PR checklist acknowledges this.

Medium: No documentation updates

The new endpoints are not documented. Per CLAUDE.md, API changes should be reflected in docs. If these endpoints are intended for external use, a docs page and sitemap/mod.ts update are needed.

Low: stopActor error parameter is undocumented and untested

engine/sdks/typescript/envoy-client/src/handle.ts adds an optional error?: string parameter to stopActor. It is unclear what the engine does with this field and there is no documentation or test coverage.

---

What Looks Good

• Type centralization: Moving kv_get types into api-types and sharing them across api-public and api-peer is a clean improvement that reduces duplication.
• Authorization pattern: Both reschedule.rs and sleep.rs correctly verify the actor belongs to the caller's namespace before sending signals, using pegboard::errors::Actor::NotFound to avoid leaking cross-namespace actor existence.
• Defensive signal sending: Using .graceful_not_found() and logging a warning when the workflow is already gone is good defensive coding.
• "stopping" SSE design: The mechanism is non-breaking since unknown SSE events were already warned-and-ignored before this change.
• ShutdownReason type: Exporting a discriminated type instead of a boolean is a better API design.
• Test envoy cleanup: Removing the dead /foo route is good housekeeping.

---

Overall

The signal rename is the most important issue to resolve — it should either follow the versioning protocol (new signal name + compat bridge) or be confirmed safe if actors with the old signal cannot survive a deployment boundary. The missing await on writeSSE is a straightforward functional bug. Tests and docs can follow in a subsequent PR if the feature is behind a flag, but both of those issues should be addressed before merge.

[NathanFlurry]

Merge activity

• Apr 5, 11:11 AM UTC: A user started a stack merge that includes this pull request via Graphite.
• Apr 5, 11:41 AM UTC: Graphite rebased this pull request as part of a merge.
• Apr 5, 11:41 AM UTC: @NathanFlurry merged this pull request with Graphite.