stream responses and all tests passing

tests/test_sandbox.py

@@ -67,11 +67,13 @@ async def test_sandbox_run_executes_command():
         sb = PodmanSandbox()
         await sb.__aenter__()
 
-        result = await sb.run("echo 'hello from sandbox'")
+        result = sb.run("echo 'hello from sandbox'")
 
         # Verify exec_run was called with shell wrapper
         mock_container.exec_run.assert_called_once_with(
-            ["/bin/sh", "-c", "echo 'hello from sandbox'"], workdir="/workspace"
+            ["/bin/sh", "-c", "echo 'hello from sandbox'"],
+            workdir="/workspace",
+            demux=False,
         )
         assert result == "hello from sandbox\n"
 
@@ -83,7 +85,7 @@ async def test_tool_call_fails_if_sandbox_crashes():
 
     # Simulate crashed sandbox (container is None)
     mock_sandbox = MagicMock()
-    mock_sandbox.run = AsyncMock(side_effect=RuntimeError("Container crashed"))
+    mock_sandbox.run = MagicMock(side_effect=RuntimeError("Container crashed"))
 
     result = await bash("ls -la", mock_sandbox)