Merge branch 'fix/ipc-24-25-codegen'

# Conflicts:
#	archreview/2026-07-12/remediation/00-tracking.md

clients/rust/protos/mxaccess_gateway.proto

@@ -676,6 +676,10 @@ message WorkerInfoReply {
 }
 
 message DrainEventsReply {
+  // The reply is bounded by both a server-side count cap and the negotiated
+  // worker-frame byte cap; a reply may therefore carry fewer events than
+  // `max_events` and fewer than are queued. Callers drain iteratively until an
+  // empty reply.
   repeated MxEvent events = 1;
 }
 
@@ -760,6 +764,11 @@ message ReplayGap {
   // after_worker_sequence = oldest_available_sequence - 1 in the next
   // StreamEventsRequest, which will cause the server to replay starting at
   // oldest_available_sequence (the first retained event).
+  // When nothing is retained (the replay ring is empty), this is the next sequence
+  // that can be delivered — `highest observed + 1` — and the `oldest - 1` resume
+  // formula remains valid: it resolves to the highest sequence already seen, so the
+  // follow-up resume replays nothing, reports no gap, and every newer live event
+  // passes. The interval evicted is unchanged.
   uint64 oldest_available_sequence = 2;
 }
 

clients/rust/protos/mxaccess_worker.proto

@@ -47,6 +47,9 @@ message GatewayHello {
   // instead of a hard-coded default; 0 (an older gateway that never set the field) means
   // "use the worker's built-in default". Sits above the public gRPC cap by an
   // envelope-overhead margin so an accepted gRPC payload always fits one worker frame.
+  // Every worker->gateway frame — events, heartbeats, faults, and control replies
+  // including DrainEvents — must serialize within this limit; reply builders truncate
+  // to fit rather than emit an oversized frame.
   uint32 max_frame_bytes = 4;
 }