chore(envoy-client): log command and event payloads

[NathanFlurry]

Summary

• log each decoded envoy command wrapper at info level when commands arrive
• log each event wrapper at info level when events are sent or resent

Verification

• cargo check -p rivet-envoy-client

[claude]

PR Review: chore(envoy-client): log command and event payloads

PR #4751 — Draft | adds per-command and per-event tracing::info! log lines using the existing stringify.rs module.

---

Overview

Adds diagnostic logging to handle_commands, handle_send_events, and resend_unacknowledged_events so each individual protocol message is logged. Builds on top of the already-existing stringify.rs module.

---

Issues

Log level is too high for per-item messages

All new lines use tracing::info!. In any environment with real traffic, this will emit one log line per command and one per event — on top of the existing batch-count lines that already exist at info. This will flood logs in production and increase storage costs significantly.

Recommendation: use tracing::debug! (or tracing::trace!) for the per-item lines. The batch-count lines (received commands, sending events, resending unacknowledged events) are appropriate at info.

// Keep count at info:
tracing::info!(command_count = commands.len(), "received commands");
// Move per-item to debug:
for command_wrapper in &commands {
    tracing::debug!(
        command = %stringify_command_wrapper(command_wrapper),
        "received command"
    );
}

Eager string allocation regardless of log level

stringify_command_wrapper and stringify_event_wrapper return owned Strings and are called unconditionally — the tracing macro evaluates all field expressions before checking whether the level is enabled. In a production binary with info as the minimum level, moving to debug! would let the subscriber short-circuit before the strings are built. If these must stay at info, guard with:

if tracing::enabled!(tracing::Level::INFO) {
    tracing::info!(command = %stringify_command_wrapper(w), "received command");
}

actor_id and generation should be separate structured fields

The CLAUDE.md convention is to use discrete structured fields (e.g. actor_id = %..., generation = ...) rather than embedding them inside a formatted string. Separate fields make it possible to filter/query logs by actor in structured log sinks.

tracing::debug!(
    actor_id = %wrapper.checkpoint.actor_id,
    generation = wrapper.checkpoint.generation,
    index = wrapper.checkpoint.index,
    command = %stringify_command(&wrapper.inner),
    "received command"
);

Extra iteration in handle_commands

The logging loop iterates over &commands and then the processing loop iterates again. This is harmless in practice (commands are small batches), but logging inline during processing would avoid the double pass.

---

Minor

• events.rs adds a new tracing::info!(event_count = ..., "sending events") line that didn't exist before — good addition at the right level.
• The stringify.rs module itself is well-structured and handles all protocol variants cleanly.

---

Summary

The stringify.rs foundation is solid. The main ask before merging: drop the per-item log lines from info to debug, which fixes both the production noise concern and the eager-allocation concern in one change.

[NathanFlurry]

• chore(kitchen-sink): configurable endpoint #4766 : 3 dependent PRs (#4754 , #4777 , #4778 )
• refactor(rivetkit-core): rename StopReason to ShutdownKind and kill enum fallthroughs #4765
• chore(envoy-client): log command and event payloads #4751 👈 (View in Graphite)
• 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.

[github-actions]

Preview packages published to npm

Install with:

npm install rivetkit@pr-4751

All packages published as 0.0.0-pr.4751.f6ab1d3 with tag pr-4751.

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-f6ab1d3
docker pull rivetdev/engine:full-f6ab1d3

Individual packages

npm install rivetkit@pr-4751
npm install @rivetkit/react@pr-4751
npm install @rivetkit/rivetkit-napi@pr-4751
npm install @rivetkit/workflow-engine@pr-4751