Improve task resilience and LLM retry visibility

Persist and publish recoverable provider retries so temporary 503 and throttling events remain visible in live and historical task streams.

Aggregate repeated retries into an expandable action card, preserve terminal pause recovery, harden Windows worker isolation and output forwarding, and retain reliable Android hierarchy fallbacks with regression coverage.

tests/unit/test_llm_pause.py

@@ -1,8 +1,11 @@
+import logging
+
 from artemis.services import llm
 
 
 class _RecordingEngine:
     current_step_id = "step-7"
+    current_session_id = "session-1"
 
     def __init__(self):
         self.trace_kwargs = None
@@ -10,6 +13,7 @@ class _RecordingEngine:
 
     def record_trace(self, **kwargs):
         self.trace_kwargs = kwargs
+        return "trace-1"
 
     def _publish(self, event_type, payload):
         self.published = (event_type, payload)
@@ -35,3 +39,42 @@ def test_pause_error_is_persisted_and_published(monkeypatch, tmp_path):
     assert payload["error"] == "503 high demand"
     assert payload["step_id"] == "step-7"
     assert isinstance(payload["timestamp"], float)
+
+
+def test_provider_sdk_retry_is_persisted_and_published(monkeypatch):
+    engine = _RecordingEngine()
+    monkeypatch.setattr(llm, "_CURRENT_DATA_ENGINE", engine)
+    record = logging.LogRecord(
+        name="google_genai._api_client",
+        level=logging.INFO,
+        pathname=__file__,
+        lineno=1,
+        msg=(
+            "Retrying google.genai request in 1.18 seconds as it raised "
+            "ServerError: 503 UNAVAILABLE: high demand"
+        ),
+        args=(),
+        exc_info=None,
+    )
+
+    llm._ProviderRetryTelemetryHandler().emit(record)
+
+    assert engine.trace_kwargs == {
+        "type": "llm_call",
+        "name": "llm_retry",
+        "payload": {
+            "error": "ServerError: 503 UNAVAILABLE: high demand",
+            "delay": 1.18,
+            "provider": "google",
+            "source": "provider_sdk",
+            "recoverable": True,
+        },
+        "step_id": "step-7",
+        "parent_trace_id": None,
+        "status": "retrying",
+    }
+    event_type, payload = engine.published
+    assert event_type == "llm_retrying"
+    assert payload["trace_id"] == "trace-1"
+    assert payload["delay"] == 1.18
+    assert payload["error"].startswith("ServerError: 503")