fix(agent): restore safe non-streaming fallback after stream failures (#3020)

After streaming retries are exhausted on transient errors, fall back to
non-streaming instead of propagating the error. Also fall back for any
other pre-delivery stream error (not just 'streaming not supported').

Added user-facing message when streaming is not supported by a model/
provider, directing users to set display.streaming: false in config.yaml
to avoid the fallback delay.

Cherry-picked from PR #3008 by kshitijk4poor. Added UX message for
streaming-not-supported detection.

Co-authored-by: kshitijk4poor <kshitijk4poor@users.noreply.github.com>

tests/test_streaming.py

@@ -487,6 +487,51 @@ class TestStreamingFallback:
         with pytest.raises(Exception, match="Rate limit exceeded"):
             agent._interruptible_streaming_api_call({})
 
+    @patch("run_agent.AIAgent._interruptible_api_call")
+    @patch("run_agent.AIAgent._create_request_openai_client")
+    @patch("run_agent.AIAgent._close_request_openai_client")
+    def test_exhausted_transient_stream_error_falls_back(self, mock_close, mock_create, mock_non_stream):
+        """Transient stream errors retry first, then fall back after retries are exhausted."""
+        from run_agent import AIAgent
+        import httpx
+
+        mock_client = MagicMock()
+        mock_client.chat.completions.create.side_effect = httpx.ConnectError("socket closed")
+        mock_create.return_value = mock_client
+
+        fallback_response = SimpleNamespace(
+            id="fallback",
+            model="test",
+            choices=[SimpleNamespace(
+                index=0,
+                message=SimpleNamespace(
+                    role="assistant",
+                    content="fallback after retries exhausted",
+                    tool_calls=None,
+                    reasoning_content=None,
+                ),
+                finish_reason="stop",
+            )],
+            usage=None,
+        )
+        mock_non_stream.return_value = fallback_response
+
+        agent = AIAgent(
+            model="test/model",
+            quiet_mode=True,
+            skip_context_files=True,
+            skip_memory=True,
+        )
+        agent.api_mode = "chat_completions"
+        agent._interrupt_requested = False
+
+        response = agent._interruptible_streaming_api_call({})
+
+        assert response.choices[0].message.content == "fallback after retries exhausted"
+        assert mock_client.chat.completions.create.call_count == 3
+        mock_non_stream.assert_called_once()
+        assert mock_close.call_count >= 1
+
 
 # ── Test: Reasoning Streaming ────────────────────────────────────────────