fix(lux): the reported version is the one that is running

lux-operator is scaled to 0, so the LuxNetwork CR's image tag is only a
declaration and can be arbitrarily stale. Mainnet's CR says v1.34.0 — a tag
that does not exist in the registry — while all five nodes answer
info.getNodeVersion with luxd/1.36.2; testnet's says v1.32.12 against a
running 1.36.24.

Reporting the CR tag as `version` would put a fabricated number on the
dashboard. `version` is now the observed value and `image` stays as declared,
so a drift between intent and reality is visible instead of hidden.

Co-authored-by: Hanzo Dev <[email protected]>

api/bootnode/core/chains/lux_fleet.py

@@ -437,14 +437,15 @@ async def _collect_network(
         _probe_node_rpc(http, _pod_proxy(ns, n.pod_name, http_port), n) for n in nodes
     ])
 
-    # Without a CR, derive the declared shape from what is actually running.
+    # The CR declares intent; the nodes report what is running. When they
+    # disagree — an unreconciled CR — the running version is the true one, so
+    # ``version`` is observed and ``image`` stays as declared.
     if not fleet.validators:
         fleet.validators = len(nodes)
-    if not fleet.version:
-        ver = next((n.version for n in nodes if n.version), None)
-        if ver:
-            fleet.version = ver
-            fleet.image = ver
+    observed = next((n.version for n in nodes if n.version), None)
+    if observed:
+        fleet.version = observed
+        fleet.image = fleet.image or observed
     if fleet.phase == "Unknown" and any(n.rpc_reachable for n in nodes):
         fleet.phase = "Running"
 

api/tests/test_lux_fleet_overview.py

@@ -116,7 +116,8 @@ class TestCollectNetwork:
 
         assert fleet.network == "devnet"
         assert fleet.namespace == "lux-devnet"
-        assert fleet.version == "v1.36.25"          # from CR
+        assert fleet.version == "luxd/1.36.25"      # observed, not the CR tag
+        assert fleet.image == "ghcr.io/luxfi/node:v1.36.25"   # declared
         assert fleet.plugin_set == "s3://lux-plugins-devnet/v1.3.1/"
         assert fleet.validators == 5
         assert fleet.nodes_total == 5
@@ -152,6 +153,32 @@ class TestCollectNetwork:
         assert fleet.phase == "Running"
         assert fleet.total_peers == 7
 
+    @pytest.mark.asyncio
+    async def test_running_version_wins_over_a_stale_cr(self):
+        """lux-operator is scaled to 0, so a CR tag can be arbitrarily stale.
+        Mainnet's CR says v1.34.0 — a tag that does not exist — while all five
+        nodes run luxd/1.36.2. The dashboard must show what is running."""
+        pods = {i: {"phase": "Running", "ready": True, "ip": f"10.2.0.{i}"}
+                for i in range(5)}
+        stale_cr = {"spec": {"validators": 5,
+                             "image": {"repository": "ghcr.io/luxfi/node",
+                                       "tag": "v1.34.0"}},
+                    "status": {"phase": "Running"}}
+
+        async def fake_probe(http, base, node: LuxNode):
+            node.rpc_reachable = True
+            node.version = "luxd/1.36.2"
+
+        with patch.object(lux_fleet, "_read_luxnetwork_cr", return_value=stale_cr), \
+             patch.object(lux_fleet, "_list_luxd_pods", return_value=pods), \
+             patch.object(lux_fleet, "_probe_node_rpc", side_effect=fake_probe), \
+             patch.object(lux_fleet, "_dex_summary", side_effect=lambda n, net: _dex_stub()):
+            fleet = await lux_fleet._collect_network(
+                "mainnet", custom_api=object(), core_api=object(), http=object())
+
+        assert fleet.version == "luxd/1.36.2"
+        assert fleet.image == "ghcr.io/luxfi/node:v1.34.0"
+
     @pytest.mark.asyncio
     async def test_unreachable_cluster_reports_error_not_silence(self):
         def boom(*a, **kw):