diff --git a/hypergraph/hypergraph.go b/hypergraph/hypergraph.go
index 07bbd78..0ff615c 100644
--- a/hypergraph/hypergraph.go
+++ b/hypergraph/hypergraph.go
@@ -132,6 +132,9 @@ func (hg *HypergraphCRDT) SetSelfPeerID(peerID string) {
 
 func (hg *HypergraphCRDT) SetShutdownContext(ctx context.Context) {
 	hg.shutdownCtx = ctx
+	// Reopen the snapshot manager in case it was closed by a previous
+	// shutdown context (i.e. during in-process engine respawn).
+	hg.snapshotMgr.reopen()
 	go func() {
 		select {
 		case <-hg.shutdownCtx.Done():
diff --git a/hypergraph/snapshot_manager.go b/hypergraph/snapshot_manager.go
index 3f0c2ac..ed63a07 100644
--- a/hypergraph/snapshot_manager.go
+++ b/hypergraph/snapshot_manager.go
@@ -462,6 +462,21 @@ func (m *snapshotManager) close() {
 	m.logger.Debug("snapshot manager closed")
 }
 
+// reopen resets the closed flag so the snapshot manager can accept new
+// snapshots after a respawn. Any previously held snapshots were already
+// released by close(), so we start with an empty generation list.
+func (m *snapshotManager) reopen() {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	if !m.closed {
+		return
+	}
+	m.closed = false
+	m.generations = make([]*snapshotGeneration, 0, maxSnapshotGenerations)
+	m.logger.Debug("snapshot manager reopened")
+}
+
 func shardKeyString(sk tries.ShardKey) string {
 	buf := make([]byte, 0, len(sk.L1)+len(sk.L2))
 	buf = append(buf, sk.L1[:]...)
diff --git a/node/consensus/app/app_consensus_engine.go b/node/consensus/app/app_consensus_engine.go
index d2c2558..e14f217 100644
--- a/node/consensus/app/app_consensus_engine.go
+++ b/node/consensus/app/app_consensus_engine.go
@@ -984,6 +984,8 @@ func (e *AppConsensusEngine) Stop(force bool) <-chan error {
 	e.pubsub.UnregisterValidator(e.getFrameMessageBitmask())
 	e.pubsub.Unsubscribe(e.getGlobalFrameMessageBitmask(), false)
 	e.pubsub.UnregisterValidator(e.getGlobalFrameMessageBitmask())
+	e.pubsub.Unsubscribe(e.getGlobalProverMessageBitmask(), false)
+	e.pubsub.UnregisterValidator(e.getGlobalProverMessageBitmask())
 	e.pubsub.Unsubscribe(e.getGlobalAlertMessageBitmask(), false)
 	e.pubsub.UnregisterValidator(e.getGlobalAlertMessageBitmask())
 	e.pubsub.Unsubscribe(e.getGlobalPeerInfoMessageBitmask(), false)
diff --git a/node/datarpc/data_worker_ipc_server.go b/node/datarpc/data_worker_ipc_server.go
index 183d2dd..3e2d69b 100644
--- a/node/datarpc/data_worker_ipc_server.go
+++ b/node/datarpc/data_worker_ipc_server.go
@@ -180,16 +180,13 @@ func (r *DataWorkerIPCServer) Respawn(
 
 func (r *DataWorkerIPCServer) RespawnServer(filter []byte) error {
 	if r.appConsensusEngine != nil {
-		// Re-respawn: gracefully shut down the process for a clean restart.
-		// The master's spawn loop (manager.go) detects the exit and
-		// immediately restarts the worker process, giving it fresh
-		// memory, a clean pubsub mesh, and a ProverRegistry built
-		// from the current on-disk hypergraph state.
-		r.logger.Info("re-respawn requested, shutting down worker for clean restart",
-			zap.Uint32("core_id", r.coreId),
-		)
-		r.Stop()
-		return nil
+		r.logger.Info("respawning worker: stopping old engine")
+		if r.cancel != nil {
+			r.cancel()
+		}
+		<-r.appConsensusEngine.Stop(false)
+		r.appConsensusEngine = nil
+		r.logger.Info("respawning worker: old engine stopped")
 	}
 	if r.server != nil {
 		r.logger.Info("stopping server for respawn")