feat: wire running service details to real status

cmd/aima/tooldeps_deploy.go

@@ -693,6 +693,7 @@ type deploymentOverview struct {
 	Name                string `json:"name"`
 	Model               string `json:"model"`
 	Engine              string `json:"engine,omitempty"`
+	Image               string `json:"image,omitempty"`
 	Slot                string `json:"slot,omitempty"`
 	Phase               string `json:"phase"`
 	Status              string `json:"status"`
@@ -704,6 +705,8 @@ type deploymentOverview struct {
 	Message             string `json:"message,omitempty"`
 	Restarts            int    `json:"restarts,omitempty"`
 	ExitCode            *int   `json:"exit_code,omitempty"`
+	GPUMemoryMiB        int    `json:"gpu_memory_mib,omitempty"`
+	GPUMemorySource     string `json:"gpu_memory_source,omitempty"`
 	StartupPhase        string `json:"startup_phase,omitempty"`
 	StartupProgress     int    `json:"startup_progress,omitempty"`
 	StartupMessage      string `json:"startup_message,omitempty"`
@@ -724,6 +727,7 @@ func deploymentOverviewFromStatus(status *runtime.DeploymentStatus, cat *knowled
 		Name:                status.Name,
 		Model:               status.Model,
 		Engine:              status.Engine,
+		Image:               status.Image,
 		Slot:                status.Slot,
 		Phase:               status.Phase,
 		Status:              status.Phase,
@@ -735,6 +739,8 @@ func deploymentOverviewFromStatus(status *runtime.DeploymentStatus, cat *knowled
 		Message:             status.Message,
 		Restarts:            status.Restarts,
 		ExitCode:            status.ExitCode,
+		GPUMemoryMiB:        status.GPUMemoryMiB,
+		GPUMemorySource:     status.GPUMemorySource,
 		StartupPhase:        status.StartupPhase,
 		StartupProgress:     status.StartupProgress,
 		StartupMessage:      status.StartupMessage,

cmd/aima/tooldeps_deploy_test.go

@@ -36,12 +36,21 @@ func TestDeploymentOverviewIncludesCatalogModelType(t *testing.T) {
 		}},
 	}
 	overview := deploymentOverviewFromStatus(&runtime.DeploymentStatus{
-		Name:  "qwen3-tts-0.6b-qwen-tts-fastapi",
-		Model: "qwen3-tts-0.6b",
-		Phase: "running",
-		Ready: true,
+		Name:            "qwen3-tts-0.6b-qwen-tts-fastapi",
+		Model:           "qwen3-tts-0.6b",
+		Image:           "docker.1ms.run/example/qwen-tts:latest",
+		Phase:           "running",
+		Ready:           true,
+		GPUMemoryMiB:    1536,
+		GPUMemorySource: "nvidia-smi",
 	}, cat)
 	if overview.ModelType != "tts" {
 		t.Fatalf("ModelType = %q, want tts", overview.ModelType)
 	}
+	if overview.Image != "docker.1ms.run/example/qwen-tts:latest" {
+		t.Fatalf("Image = %q, want docker.1ms.run/example/qwen-tts:latest", overview.Image)
+	}
+	if overview.GPUMemoryMiB != 1536 || overview.GPUMemorySource != "nvidia-smi" {
+		t.Fatalf("GPU memory = %d/%q, want 1536/nvidia-smi", overview.GPUMemoryMiB, overview.GPUMemorySource)
+	}
 }

internal/k3s/client.go

@@ -41,6 +41,7 @@ type PodStatus struct {
 	DeletionTimestamp string            `json:"deletion_timestamp,omitempty"`
 	Message           string            `json:"message,omitempty"`
 	ContainerPort     int               `json:"container_port,omitempty"`
+	ContainerImage    string            `json:"container_image,omitempty"`
 	RestartCount      int               `json:"restart_count,omitempty"`
 	ExitCode          *int              `json:"exit_code,omitempty"`         // from Terminated state
 	ContainerStarted  string            `json:"container_started,omitempty"` // when the current container instance started
@@ -270,6 +271,7 @@ type kubePod struct {
 	} `json:"metadata"`
 	Spec struct {
 		Containers []struct {
+			Image string `json:"image"`
 			Ports []struct {
 				ContainerPort int `json:"containerPort"`
 			} `json:"ports"`
@@ -357,8 +359,12 @@ func parsePodJSON(data []byte) (*PodStatus, error) {
 	}
 
 	containerPort := 0
+	containerImage := ""
 	if len(kp.Spec.Containers) > 0 && len(kp.Spec.Containers[0].Ports) > 0 {
+		containerImage = kp.Spec.Containers[0].Image
 		containerPort = kp.Spec.Containers[0].Ports[0].ContainerPort
+	} else if len(kp.Spec.Containers) > 0 {
+		containerImage = kp.Spec.Containers[0].Image
 	}
 
 	var conditions []PodCondition
@@ -376,6 +382,7 @@ func parsePodJSON(data []byte) (*PodStatus, error) {
 		DeletionTimestamp: kp.Metadata.DeletionTimestamp,
 		Message:           msg,
 		ContainerPort:     containerPort,
+		ContainerImage:    containerImage,
 		RestartCount:      restartCount,
 		ExitCode:          exitCode,
 		ContainerStarted:  containerStarted,

internal/k3s/client_test.go

@@ -195,6 +195,18 @@ const terminatingPodJSON = `{
       "aima.dev/model": "qwen3-8b"
     }
   },
+  "spec": {
+    "containers": [
+      {
+        "image": "nvcr.io/nvidia/vllm:26.01-py3",
+        "ports": [
+          {
+            "containerPort": 8000
+          }
+        ]
+      }
+    ]
+  },
   "status": {
     "phase": "Running",
     "podIP": "10.42.0.5",
@@ -217,6 +229,18 @@ const runningPodJSON = `{
       "aima.dev/model": "qwen3-8b"
     }
   },
+  "spec": {
+    "containers": [
+      {
+        "image": "nvcr.io/nvidia/vllm:26.01-py3",
+        "ports": [
+          {
+            "containerPort": 8000
+          }
+        ]
+      }
+    ]
+  },
   "status": {
     "phase": "Running",
     "podIP": "10.42.0.5",
@@ -621,3 +645,16 @@ func TestParsePodJSON_DeletionTimestamp(t *testing.T) {
 		t.Fatal("expected raw pod readiness to reflect container status before runtime mapping")
 	}
 }
+
+func TestParsePodJSON_ContainerImage(t *testing.T) {
+	pod, err := parsePodJSON([]byte(runningPodJSON))
+	if err != nil {
+		t.Fatal(err)
+	}
+	if pod.ContainerImage != "nvcr.io/nvidia/vllm:26.01-py3" {
+		t.Fatalf("ContainerImage = %q, want nvcr.io/nvidia/vllm:26.01-py3", pod.ContainerImage)
+	}
+	if pod.ContainerPort != 8000 {
+		t.Fatalf("ContainerPort = %d, want 8000", pod.ContainerPort)
+	}
+}