feat: add runtime agent version tracking and update metrics reporting

Author: goktugcy

internal/agent/service.go

@@ -13,8 +13,8 @@ import (
 	"github.com/noderax/noderax-agent/internal/config"
 	"github.com/noderax/noderax-agent/internal/metrics"
 	"github.com/noderax/noderax-agent/internal/realtime"
-	"github.com/noderax/noderax-agent/internal/terminal"
 	"github.com/noderax/noderax-agent/internal/tasks"
+	"github.com/noderax/noderax-agent/internal/terminal"
 )
 
 type Service struct {
@@ -148,6 +148,9 @@ func NewService(cfg config.Config, client *api.Client, logger *slog.Logger, vers
 	if err != nil {
 		logger.Error("failed to initialize realtime service", "error", err)
 	}
+	if realtimeService != nil {
+		realtimeService.SetRuntimeAgentVersion(version)
+	}
 
 	var terminalEvents terminal.RealtimeEvents
 	if realtimeService != nil {

internal/realtime/client.go

@@ -44,18 +44,19 @@ type Stats struct {
 }
 
 type Service struct {
-	logger         *slog.Logger
-	requestTimeout time.Duration
-	pingInterval   time.Duration
-	jitterRatio    float64
-	credentials    func() (string, string)
-	dispatcher     *dispatcher
-	onAuthSuccess  AuthSuccessHook
-	dialURL        string
-	healthURL      string
-	namespace      string
-	path           string
-	outbound       chan any
+	logger              *slog.Logger
+	requestTimeout      time.Duration
+	pingInterval        time.Duration
+	jitterRatio         float64
+	credentials         func() (string, string)
+	dispatcher          *dispatcher
+	onAuthSuccess       AuthSuccessHook
+	dialURL             string
+	healthURL           string
+	namespace           string
+	path                string
+	runtimeAgentVersion string
+	outbound            chan any
 
 	reconnects      atomic.Int64
 	pingsSent       atomic.Int64
@@ -352,6 +353,7 @@ func (s *Service) SendMetrics(ctx context.Context, event api.MetricsRequest) err
 		Type:         EventAgentMetrics,
 		NodeID:       event.NodeID,
 		AgentToken:   event.AgentToken,
+		AgentVersion: s.runtimeAgentVersion,
 		Timestamp:    formatTimestampUTCMillis(event.CollectedAt),
 		CPUUsage:     &cpuUsage,
 		MemoryUsage:  &memoryUsage,
@@ -374,6 +376,10 @@ func (s *Service) ReportDispatchHandled() {
 	s.dispatchHandled.Add(1)
 }
 
+func (s *Service) SetRuntimeAgentVersion(version string) {
+	s.runtimeAgentVersion = strings.TrimSpace(version)
+}
+
 func (s *Service) SnapshotStats() Stats {
 	return Stats{
 		Reconnects:      s.reconnects.Load(),
@@ -512,7 +518,12 @@ func (s *Service) connect(ctx context.Context) (*socketIOConn, error) {
 	}
 
 	socket.OnConnect(func() {
-		authPayload := authEvent{Type: EventAgentAuth, NodeID: nodeID, AgentToken: agentToken}
+		authPayload := authEvent{
+			Type:         EventAgentAuth,
+			NodeID:       nodeID,
+			AgentToken:   agentToken,
+			AgentVersion: s.runtimeAgentVersion,
+		}
 		socket.Emit(EventAgentAuth, authPayload)
 	})
 

internal/realtime/client_test.go

@@ -57,6 +57,7 @@ func TestSendMetricsUsesMillisecondTimestamp(t *testing.T) {
 
 	svc := &Service{outbound: make(chan any, 1), logger: slog.New(slog.NewTextHandler(io.Discard, nil))}
 	svc.sessionActive.Store(true)
+	svc.SetRuntimeAgentVersion("1.0.0")
 	collectedAt := time.Date(2026, 3, 20, 10, 20, 30, 456789123, time.UTC)
 
 	err := svc.SendMetrics(context.Background(), api.MetricsRequest{
@@ -82,6 +83,9 @@ func TestSendMetricsUsesMillisecondTimestamp(t *testing.T) {
 	if event.AgentToken != "token-1" {
 		t.Fatalf("unexpected agent token: %q", event.AgentToken)
 	}
+	if event.AgentVersion != "1.0.0" {
+		t.Fatalf("unexpected agent version: %q", event.AgentVersion)
+	}
 
 	pattern := regexp.MustCompile(`^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}\.\d{3}Z$`)
 	if !pattern.MatchString(event.Timestamp) {

internal/realtime/handlers.go

@@ -10,12 +10,12 @@ import (
 )
 
 const (
-	EventAgentAuth    = "agent.auth"
-	EventAgentAuthAck = "agent.auth.ack"
-	EventAgentAuthErr = "agent.auth.error"
-	EventAgentError   = "agent.error"
-	EventAgentPing    = "agent.ping"
-	EventAgentMetrics = "agent.metrics"
+	EventAgentAuth      = "agent.auth"
+	EventAgentAuthAck   = "agent.auth.ack"
+	EventAgentAuthErr   = "agent.auth.error"
+	EventAgentError     = "agent.error"
+	EventAgentPing      = "agent.ping"
+	EventAgentMetrics   = "agent.metrics"
 	EventTerminalStart  = "terminal.start"
 	EventTerminalInput  = "terminal.input"
 	EventTerminalResize = "terminal.resize"
@@ -24,11 +24,11 @@ const (
 	EventTerminalOutput = "terminal.output"
 	EventTerminalExited = "terminal.exited"
 	EventTerminalError  = "terminal.error"
-	EventTaskDispatch = "task.dispatch"
-	EventTaskAccepted = "task.accepted"
-	EventTaskStarted  = "task.started"
-	EventTaskLog      = "task.log"
-	EventTaskComplete = "task.completed"
+	EventTaskDispatch   = "task.dispatch"
+	EventTaskAccepted   = "task.accepted"
+	EventTaskStarted    = "task.started"
+	EventTaskLog        = "task.log"
+	EventTaskComplete   = "task.completed"
 )
 
 type taskDispatcher interface {
@@ -128,9 +128,10 @@ func (d *dispatcher) handleMessage(ctx context.Context, data []byte) error {
 }
 
 type authEvent struct {
-	Type       string `json:"type"`
-	NodeID     string `json:"nodeId"`
-	AgentToken string `json:"agentToken"`
+	Type         string `json:"type"`
+	NodeID       string `json:"nodeId"`
+	AgentToken   string `json:"agentToken"`
+	AgentVersion string `json:"agentVersion,omitempty"`
 }
 
 type pingEvent struct {
@@ -142,6 +143,7 @@ type metricsEvent struct {
 	Type         string             `json:"type"`
 	NodeID       string             `json:"nodeId"`
 	AgentToken   string             `json:"agentToken"`
+	AgentVersion string             `json:"agentVersion,omitempty"`
 	Timestamp    string             `json:"timestamp"`
 	CPUUsage     *float64           `json:"cpuUsage,omitempty"`
 	MemoryUsage  *float64           `json:"memoryUsage,omitempty"`

internal/realtime/integration_test.go

@@ -139,6 +139,7 @@ func TestRealtimeConnectAuthDispatchLifecycle(t *testing.T) {
 	if err != nil {
 		t.Fatalf("NewService() error = %v", err)
 	}
+	svc.SetRuntimeAgentVersion("1.0.0")
 
 	runCtx, cancel := context.WithCancel(context.Background())
 	done := make(chan error, 1)
@@ -151,6 +152,9 @@ func TestRealtimeConnectAuthDispatchLifecycle(t *testing.T) {
 		if auth.Type != EventAgentAuth || auth.NodeID != "node-1" || auth.AgentToken != "token-1" {
 			t.Fatalf("unexpected auth payload: %+v", auth)
 		}
+		if auth.AgentVersion != "1.0.0" {
+			t.Fatalf("unexpected auth payload: %+v", auth)
+		}
 	case <-time.After(8 * time.Second):
 		t.Fatalf("timed out waiting for auth event")
 	}