fix(hermes-autotriage): pin repair Job to Longhorn-ready worker nodes

Non-worker nodes (titan-2x lane) are not Longhorn-ready; volume attach
fails there. Same constraint as the demo test-runner.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

ariadne/services/hermes_autotriage_repair.py

@@ -99,6 +99,7 @@ def _job_payload(cfg: dict, job_name: str, incident_id: str) -> dict[str, Any]:
                 "metadata": {"labels": {"app.kubernetes.io/part-of": "hermes-triage-demo"}},
                 "spec": {
                     "restartPolicy": "Never",
+                    "nodeSelector": {"node-role.kubernetes.io/worker": "true"},
                     "containers": [
                         {
                             "name": "repair",

tests/test_hermes_autotriage_repair.py

@@ -92,6 +92,7 @@ def test_repair_job_payload_contract(monkeypatch) -> None:
     assert spec["ttlSecondsAfterFinished"] == 3600
     pod = spec["template"]["spec"]
     assert pod["restartPolicy"] == "Never"
+    assert pod["nodeSelector"] == {"node-role.kubernetes.io/worker": "true"}
     container = pod["containers"][0]
     assert container["image"] == "busybox:1.37"
     assert container["command"][:2] == ["sh", "-c"]