fix: cross-platform compatibility for ScriptExecutor and tests

src/conductor/cli/run.py

@@ -72,7 +72,7 @@ def init_file_logging(log_path: Path) -> None:
     """
     global _file_console, _file_handle
     log_path.parent.mkdir(parents=True, exist_ok=True)
-    _file_handle = open(log_path, "w")  # noqa: SIM115
+    _file_handle = open(log_path, "w", encoding="utf-8")  # noqa: SIM115
     _file_console = Console(file=_file_handle, no_color=True, highlight=False, width=200)
 
 

tests/test_cli/test_logging.py

@@ -11,9 +11,11 @@
 from __future__ import annotations
 
 import contextlib
+import sys
 from pathlib import Path
 from unittest.mock import patch
 
+import pytest
 from typer.testing import CliRunner
 
 from conductor.cli.app import (
@@ -732,7 +734,7 @@ def test_verbose_log_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log("test file message")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "test file message" in content
         finally:
             verbose_mode.reset(token)
@@ -750,7 +752,7 @@ def test_file_output_is_plain_text(self, tmp_path: Path) -> None:
             verbose_log("styled message", style="bold red")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "styled message" in content
             # No ANSI escape sequences
             ansi_pattern = re.compile(r"\x1b\[[0-9;]*m")
@@ -772,7 +774,7 @@ def test_file_gets_untruncated_content(self, tmp_path: Path) -> None:
                 verbose_log_section("Test", long_content)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             # File should have full content, not truncated
             assert "truncated" not in content
             assert content.count("x") == 1000
@@ -797,7 +799,7 @@ def test_file_logging_in_silent_mode(self, tmp_path: Path) -> None:
             verbose_log_timing("test op", 1.5)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "silent mode message" in content
             assert "test op" in content
             assert "1.50" in content
@@ -1023,7 +1025,7 @@ def test_file_gets_sections_in_minimal_mode(self, tmp_path: Path) -> None:
                 verbose_log_section("Prompt", "full prompt content here")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "Prompt" in content
             assert "full prompt content here" in content
         finally:
@@ -1078,7 +1080,7 @@ def test_agent_start_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_agent_start("test-agent", 1)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "test-agent" in content
             assert "iter 1" in content
         finally:
@@ -1101,7 +1103,7 @@ def test_agent_complete_writes_to_file(self, tmp_path: Path) -> None:
             )
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "test-agent" in content
             assert "1.50" in content
             assert "gpt-4" in content
@@ -1120,7 +1122,7 @@ def test_route_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_route("next-agent")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "next-agent" in content
         finally:
             verbose_mode.reset(token)
@@ -1136,7 +1138,7 @@ def test_route_end_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_route("$end")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "$end" in content
         finally:
             verbose_mode.reset(token)
@@ -1152,7 +1154,7 @@ def test_timing_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_timing("Workflow execution", 3.456)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "Workflow execution" in content
             assert "3.46" in content
         finally:
@@ -1173,7 +1175,7 @@ def test_parallel_start_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_parallel_start("parallel-group", 3)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "parallel-group" in content
             assert "3 agents" in content
         finally:
@@ -1194,7 +1196,7 @@ def test_parallel_agent_complete_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_parallel_agent_complete("agent-a", 2.0, model="gpt-4", tokens=200)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "agent-a" in content
             assert "2.00" in content
         finally:
@@ -1215,7 +1217,7 @@ def test_parallel_agent_failed_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_parallel_agent_failed("agent-b", 0.5, "RuntimeError", "Something broke")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "agent-b" in content
             assert "RuntimeError" in content
             assert "Something broke" in content
@@ -1237,7 +1239,7 @@ def test_parallel_summary_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_parallel_summary("group1", 2, 1, 3.0)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "group1" in content
             assert "2 succeeded" in content
             assert "1 failed" in content
@@ -1259,7 +1261,7 @@ def test_for_each_start_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_for_each_start("loop-group", 5, 2, "fail_fast")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "loop-group" in content
             assert "5 items" in content
         finally:
@@ -1280,7 +1282,7 @@ def test_for_each_item_complete_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_for_each_item_complete("item-0", 1.2, tokens=50)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "item-0" in content
             assert "1.20" in content
         finally:
@@ -1301,7 +1303,7 @@ def test_for_each_item_failed_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_for_each_item_failed("item-2", 0.3, "ValueError", "bad input")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "item-2" in content
             assert "ValueError" in content
             assert "bad input" in content
@@ -1323,7 +1325,7 @@ def test_for_each_summary_writes_to_file(self, tmp_path: Path) -> None:
             verbose_log_for_each_summary("loop1", 4, 1, 5.0)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "loop1" in content
             assert "4 succeeded" in content
             assert "1 failed" in content
@@ -1351,7 +1353,7 @@ def test_display_usage_summary_writes_to_file(self, tmp_path: Path) -> None:
             )
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "Token Usage" in content
             assert "500" in content
             assert "200" in content
@@ -1371,7 +1373,7 @@ def test_section_writes_full_content_to_file_in_silent_mode(self, tmp_path: Path
             verbose_log_section("Prompt", "full prompt content")
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             assert "Prompt" in content
             assert "full prompt content" in content
         finally:
@@ -1528,6 +1530,7 @@ def test_log_path_not_printed_when_init_fails(self, tmp_path: Path) -> None:
 class TestFileLoggingErrorHandling:
     """Tests for file logging error handling."""
 
+    @pytest.mark.skipif(sys.platform == "win32", reason="Unix-specific chmod permissions")
     def test_init_file_logging_permission_denied(self, tmp_path: Path) -> None:
         """Test that init_file_logging raises OSError for permission issues."""
         import os
@@ -1551,6 +1554,7 @@ def test_init_file_logging_permission_denied(self, tmp_path: Path) -> None:
             # Restore permissions for cleanup
             os.chmod(readonly_dir, 0o755)
 
+    @pytest.mark.skipif(sys.platform == "win32", reason="Unix-specific path behavior")
     def test_run_workflow_handles_log_file_error_gracefully(self, tmp_path: Path) -> None:
         """Test that run_workflow_async handles log file errors gracefully."""
         import asyncio
@@ -1635,7 +1639,7 @@ def test_file_output_no_ansi_for_all_styles(self, tmp_path: Path) -> None:
             verbose_log_timing("operation", 1.5)
             close_file_logging()
 
-            content = log_path.read_text()
+            content = log_path.read_text(encoding="utf-8")
             ansi_pattern = re.compile(r"\x1b\[[0-9;]*m")
             assert not ansi_pattern.search(content), f"ANSI codes found in file output: {content}"
         finally:

tests/test_engine/test_checkpoint.py

@@ -20,7 +20,7 @@
 import stat
 import sys
 from datetime import UTC, datetime
-from pathlib import Path
+from pathlib import Path, PurePosixPath
 from typing import Any
 from unittest.mock import patch
 
@@ -106,13 +106,13 @@ def test_path(self) -> None:
         assert _make_json_serializable(p) == str(p)
 
     def test_dict_recursive(self) -> None:
-        d = {"path": Path("/a"), "nested": {"b": b"data"}}
+        d = {"path": PurePosixPath("/a"), "nested": {"b": b"data"}}
         result = _make_json_serializable(d)
         assert result["path"] == "/a"
         assert result["nested"]["b"] == "data"
 
     def test_list_recursive(self) -> None:
-        result = _make_json_serializable([Path("/a"), 42, [b"x"]])
+        result = _make_json_serializable([PurePosixPath("/a"), 42, [b"x"]])
         assert result == ["/a", 42, ["x"]]
 
     def test_set_converted_to_sorted_list(self) -> None:
@@ -244,7 +244,7 @@ def test_handles_non_serializable_inputs(self, tmp_path: Path) -> None:
         limits = _make_limits()
         error = RuntimeError("err")
 
-        inputs_with_path: dict[str, Any] = {"file": Path("/tmp/x"), "data": b"bytes"}
+        inputs_with_path: dict[str, Any] = {"file": PurePosixPath("/tmp/x"), "data": b"bytes"}
 
         with patch.object(CheckpointManager, "get_checkpoints_dir", return_value=tmp_path):
             path = CheckpointManager.save_checkpoint(wf, ctx, limits, "a", error, inputs_with_path)

tests/test_engine/test_event_emission.py

@@ -7,6 +7,7 @@
 
 from __future__ import annotations
 
+import sys
 from unittest.mock import MagicMock
 
 import pytest
@@ -510,8 +511,8 @@ async def test_script_started_and_completed(self) -> None:
                 AgentDef(
                     name="run_echo",
                     type="script",
-                    command="echo",
-                    args=["hello"],
+                    command=sys.executable,
+                    args=["-c", "print('hello')"],
                     routes=[RouteDef(to="$end")],
                 ),
             ],

tests/test_engine/test_script_workflow.py

@@ -15,6 +15,7 @@
 
 from __future__ import annotations
 
+import sys
 from unittest.mock import MagicMock
 
 import pytest
@@ -52,8 +53,8 @@ async def test_script_step_runs_to_end(self) -> None:
                 AgentDef(
                     name="run_echo",
                     type="script",
-                    command="echo",
-                    args=["hello world"],
+                    command=sys.executable,
+                    args=["-c", "print('hello world')"],
                     routes=[RouteDef(to="$end")],
                 ),
             ],
@@ -83,8 +84,8 @@ async def test_script_output_in_context(self) -> None:
                 AgentDef(
                     name="checker",
                     type="script",
-                    command="echo",
-                    args=["test output"],
+                    command=sys.executable,
+                    args=["-c", "print('test output')"],
                     routes=[RouteDef(to="processor")],
                 ),
                 AgentDef(
@@ -131,7 +132,8 @@ async def test_route_on_exit_code_simpleeval_success(self) -> None:
                 AgentDef(
                     name="checker",
                     type="script",
-                    command="true",
+                    command=sys.executable,
+                    args=["-c", "import sys; sys.exit(0)"],
                     routes=[
                         RouteDef(to="success_handler", when="exit_code == 0"),
                         RouteDef(to="failure_handler"),
@@ -175,7 +177,8 @@ async def test_route_on_exit_code_simpleeval_failure(self) -> None:
                 AgentDef(
                     name="checker",
                     type="script",
-                    command="false",
+                    command=sys.executable,
+                    args=["-c", "import sys; sys.exit(1)"],
                     routes=[
                         RouteDef(to="success_handler", when="exit_code == 0"),
                         RouteDef(to="failure_handler"),
@@ -219,7 +222,8 @@ async def test_route_on_exit_code_jinja2(self) -> None:
                 AgentDef(
                     name="checker",
                     type="script",
-                    command="true",
+                    command=sys.executable,
+                    args=["-c", "import sys; sys.exit(0)"],
                     routes=[
                         RouteDef(to="success_handler", when="{{ output.exit_code == 0 }}"),
                         RouteDef(to="failure_handler"),
@@ -267,15 +271,15 @@ async def test_script_counts_toward_iteration_limit(self) -> None:
                 AgentDef(
                     name="step1",
                     type="script",
-                    command="echo",
-                    args=["step1"],
+                    command=sys.executable,
+                    args=["-c", "print('step1')"],
                     routes=[RouteDef(to="step2")],
                 ),
                 AgentDef(
                     name="step2",
                     type="script",
-                    command="echo",
-                    args=["step2"],
+                    command=sys.executable,
+                    args=["-c", "print('step2')"],
                     routes=[RouteDef(to="$end")],
                 ),
             ],
@@ -303,7 +307,8 @@ async def test_script_non_zero_exit_no_routes_ends(self) -> None:
                 AgentDef(
                     name="failing",
                     type="script",
-                    command="false",
+                    command=sys.executable,
+                    args=["-c", "import sys; sys.exit(1)"],
                 ),
             ],
             output={
@@ -337,8 +342,8 @@ async def test_mixed_agent_and_script(self) -> None:
                 AgentDef(
                     name="setup_script",
                     type="script",
-                    command="echo",
-                    args=["setup complete"],
+                    command=sys.executable,
+                    args=["-c", "print('setup complete')"],
                     routes=[RouteDef(to="analyzer")],
                 ),
                 AgentDef(
@@ -380,8 +385,8 @@ async def test_script_command_with_workflow_input(self) -> None:
                 AgentDef(
                     name="runner",
                     type="script",
-                    command="echo",
-                    args=["{{ workflow.input.message }}"],
+                    command=sys.executable,
+                    args=["-c", "import sys; print(sys.argv[1])", "{{ workflow.input.message }}"],
                     routes=[RouteDef(to="$end")],
                 ),
             ],