fix(client/python): reachable cert-validation flag; bounded off-loop TOFU probe; license/marker fixes (Client.Python-027..031)

clients/python/README.md

@@ -238,7 +238,11 @@ left `False`, the client fetches the gateway's presented certificate once
 to `localhost` (the generated certificate always carries a `localhost` SAN) when
 none was supplied. To verify instead, pass `ca_file` to verify against a specific
 CA, or set `require_certificate_validation=True` to verify against the system
-trust roots. See
+trust roots. The strict posture is reachable through every documented entry
+point: the `require_certificate_validation=True` keyword on
+`GatewayClient.connect(...)` / `GalaxyRepositoryClient.connect(...)`, the
+`ClientOptions(require_certificate_validation=True)` struct, and the
+`--require-certificate-validation` CLI flag. See
 [Gateway Configuration](../../docs/GatewayConfiguration.md#automatic-self-signed-certificate).
 
 ## CLI
@@ -267,6 +271,13 @@ Use TLS options for a secured gateway:
 mxgw-py smoke --endpoint mxgateway.example.local:5001 --tls --ca-file C:\certs\mxgateway-ca.pem --server-name-override mxgateway.example.local --api-key-env MXGATEWAY_API_KEY --item Object.Attribute --json
 ```
 
+To force certificate validation against the system trust store instead of the
+lenient trust-on-first-use default, add `--require-certificate-validation`:
+
+```powershell
+mxgw-py smoke --endpoint mxgateway.example.local:5001 --tls --require-certificate-validation --api-key-env MXGATEWAY_API_KEY --item Object.Attribute --json
+```
+
 ## Integration Checks
 
 Run live checks only when a gateway and MXAccess-backed worker are available:

clients/python/pyproject.toml

@@ -16,11 +16,10 @@ dependencies = [
 authors = [
   { name = "Joseph Doherty" },
 ]
-license = { text = "Proprietary" }
+license = "LicenseRef-Proprietary"
 keywords = ["mxaccess", "mxgateway", "grpc", "client", "archestra"]
 classifiers = [
   "Development Status :: 3 - Alpha",
-  "License :: Other/Proprietary License",
   "Programming Language :: Python :: 3",
   "Programming Language :: Python :: 3.12",
   "Programming Language :: Python :: 3.13",
@@ -54,3 +53,6 @@ where = ["src"]
 addopts = "-ra"
 pythonpath = ["src"]
 testpaths = ["tests"]
+markers = [
+  "tls: loopback TLS tests, opt-in via MXGATEWAY_RUN_TLS_TESTS=1",
+]

clients/python/src/zb_mom_ww_mxgateway/client.py

@@ -40,6 +40,7 @@ class GatewayClient:
         api_key: str | None = None,
         plaintext: bool = False,
         ca_file: str | None = None,
+        require_certificate_validation: bool = False,
         server_name_override: str | None = None,
         stub: Any | None = None,
     ) -> "GatewayClient":
@@ -50,13 +51,16 @@ class GatewayClient:
             api_key=api_key,
             plaintext=plaintext,
             ca_file=ca_file,
+            require_certificate_validation=require_certificate_validation,
             server_name_override=server_name_override,
         )
 
         if stub is not None:
             return cls(options=resolved, stub=stub)
 
-        channel = create_channel(resolved)
+        # create_channel may perform a blocking TLS certificate probe (TOFU
+        # default); run it off the event loop so connect never freezes the loop.
+        channel = await asyncio.to_thread(create_channel, resolved)
         return cls(
             options=resolved,
             stub=pb_grpc.MxAccessGatewayStub(channel),

clients/python/src/zb_mom_ww_mxgateway/galaxy.py

@@ -52,6 +52,7 @@ class GalaxyRepositoryClient:
         api_key: str | None = None,
         plaintext: bool = False,
         ca_file: str | None = None,
+        require_certificate_validation: bool = False,
         server_name_override: str | None = None,
         stub: Any | None = None,
     ) -> "GalaxyRepositoryClient":
@@ -62,13 +63,16 @@ class GalaxyRepositoryClient:
             api_key=api_key,
             plaintext=plaintext,
             ca_file=ca_file,
+            require_certificate_validation=require_certificate_validation,
             server_name_override=server_name_override,
         )
 
         if stub is not None:
             return cls(options=resolved, stub=stub)
 
-        channel = create_channel(resolved)
+        # create_channel may perform a blocking TLS certificate probe (TOFU
+        # default); run it off the event loop so connect never freezes the loop.
+        channel = await asyncio.to_thread(create_channel, resolved)
         return cls(
             options=resolved,
             stub=galaxy_pb_grpc.GalaxyRepositoryStub(channel),

clients/python/src/zb_mom_ww_mxgateway/options.py

@@ -12,6 +12,10 @@ import grpc
 from .auth import REDACTED, ApiKey
 from .errors import MxGatewayTransportError
 
+# Fallback bound for the TOFU certificate probe when no call_timeout is set, so a
+# black-holed host fails fast instead of hanging on the OS default connect timeout.
+_TOFU_PROBE_TIMEOUT_SECONDS = 10.0
+
 
 @dataclass(frozen=True)
 class ClientOptions:
@@ -88,8 +92,17 @@ def _split_authority(endpoint: str) -> tuple[str, int]:
         remainder = target[bracket_end + 1 :]  # ":5120" or ""
         port_str = remainder.lstrip(":")
         return (host, int(port_str) if port_str else 443)
-    host, _, port = target.rpartition(":")
-    return (host or "localhost", int(port) if port else 443)
+    host, sep, port = target.rpartition(":")
+    if not sep:
+        # No colon at all (e.g. a bare hostname "mygateway"): the whole target
+        # is the host; default the port rather than raising on int("mygateway").
+        return (target or "localhost", 443)
+    if not port.isdigit():
+        # A colon with a non-numeric / empty tail (e.g. a trailing ":") is not
+        # an explicit port — keep the left side as the host and default the
+        # port so a typo cannot raise an uncaught ValueError on the TOFU path.
+        return (host or "localhost", 443)
+    return (host or "localhost", int(port))
 
 
 def create_channel(options: ClientOptions) -> grpc.aio.Channel:
@@ -120,9 +133,15 @@ def create_channel(options: ClientOptions) -> grpc.aio.Channel:
     else:
         # Lenient default: grpc-python has no per-channel skip-verify, so fetch the
         # server's certificate (unverified) and pin it for this channel (TOFU).
+        # The probe opens a real blocking TCP+TLS socket, so it MUST be bounded —
+        # a black-holed / firewall-drop host would otherwise hang on the OS default
+        # connect timeout (minutes). Bound it by call_timeout (or a short fixed
+        # fallback) so the dial fails fast as a transport error. The async
+        # `connect` classmethods run this off the event loop (asyncio.to_thread).
         host, port = _split_authority(options.endpoint)
+        probe_timeout = options.call_timeout if options.call_timeout else _TOFU_PROBE_TIMEOUT_SECONDS
         try:
-            presented = ssl.get_server_certificate((host, port))
+            presented = ssl.get_server_certificate((host, port), timeout=probe_timeout)
         except OSError as error:
             raise MxGatewayTransportError(
                 f"failed to fetch TLS certificate from {options.endpoint}: {error}"

clients/python/src/zb_mom_ww_mxgateway_cli/commands.py

@@ -170,6 +170,13 @@ def gateway_options(command: Callable[..., Any]) -> Callable[..., Any]:
     command = click.option("--plaintext", is_flag=True, help="Use plaintext gRPC.")(command)
     command = click.option("--tls", "use_tls", is_flag=True, help="Use TLS gRPC.")(command)
     command = click.option("--ca-file", default=None, help="Custom root certificate file.")(command)
+    command = click.option(
+        "--require-certificate-validation",
+        "require_certificate_validation",
+        is_flag=True,
+        help="Verify the TLS certificate against the system trust store "
+        "instead of the lenient trust-on-first-use default.",
+    )(command)
     command = click.option(
         "--server-name-override",
         default=None,
@@ -923,6 +930,7 @@ async def _connect(kwargs: dict[str, Any]) -> GatewayClient:
             api_key=api_key,
             plaintext=_use_plaintext(kwargs),
             ca_file=kwargs.get("ca_file"),
+            require_certificate_validation=bool(kwargs.get("require_certificate_validation")),
             server_name_override=kwargs.get("server_name_override"),
             call_timeout=kwargs.get("call_timeout"),
             stream_timeout=kwargs.get("stream_timeout"),

clients/python/tests/test_auth_options.py

@@ -1,9 +1,12 @@
 """Tests for auth metadata and connection options."""
 
+import socket
+
 import pytest
 
 from zb_mom_ww_mxgateway.auth import REDACTED, ApiKey, auth_metadata, redact_secret
 from zb_mom_ww_mxgateway import options as options_module
+from zb_mom_ww_mxgateway.errors import MxGatewayTransportError
 from zb_mom_ww_mxgateway.options import ClientOptions, create_channel
 
 
@@ -80,7 +83,9 @@ def test_create_channel_uses_tls_channel_tofu_default(monkeypatch: pytest.Monkey
     _DUMMY_PEM = "-----BEGIN CERTIFICATE-----\nZmFrZQ==\n-----END CERTIFICATE-----\n"
     get_cert_calls: list[tuple[str, int]] = []
 
-    def fake_get_server_certificate(addr: tuple[str, int]) -> str:
+    def fake_get_server_certificate(
+        addr: tuple[str, int], *, timeout: float | None = None
+    ) -> str:
         get_cert_calls.append(addr)
         return _DUMMY_PEM
 
@@ -133,7 +138,7 @@ def test_create_channel_uses_tls_channel_tofu_respects_server_name_override(
     monkeypatch.setattr(
         options_module.ssl,
         "get_server_certificate",
-        lambda addr: _DUMMY_PEM,
+        lambda addr, *, timeout=None: _DUMMY_PEM,
     )
     cred_calls: list[object] = []
 
@@ -276,3 +281,46 @@ def test_create_channel_uses_tls_channel_ca_file(
             ],
         ),
     ]
+
+
+def test_tofu_probe_passes_a_bounded_timeout(monkeypatch: pytest.MonkeyPatch) -> None:
+    """The TOFU cert pre-fetch must be bounded so a black-holed host fails fast."""
+    captured: dict[str, object] = {}
+
+    def fake_get_server_certificate(addr: object, *, timeout: float | None = None) -> str:
+        captured["timeout"] = timeout
+        return "-----BEGIN CERTIFICATE-----\nZmFrZQ==\n-----END CERTIFICATE-----\n"
+
+    monkeypatch.setattr(options_module.ssl, "get_server_certificate", fake_get_server_certificate)
+    monkeypatch.setattr(options_module.grpc, "ssl_channel_credentials", lambda **_: "creds")
+    monkeypatch.setattr(
+        options_module.grpc.aio,
+        "secure_channel",
+        lambda endpoint, credentials, *, options: "tls-channel",
+    )
+
+    create_channel(ClientOptions(endpoint="gateway.example:5001", call_timeout=7.5))
+
+    # A finite, positive timeout must be supplied (bounded by call_timeout here).
+    assert isinstance(captured["timeout"], (int, float))
+    assert 0 < captured["timeout"] <= 7.5
+
+
+@pytest.mark.parametrize(
+    "raised",
+    [socket.timeout("timed out"), TimeoutError("timed out"), OSError("connection refused")],
+)
+def test_tofu_probe_timeout_raises_transport_error(
+    monkeypatch: pytest.MonkeyPatch, raised: Exception
+) -> None:
+    """A timed-out / failed probe surfaces as MxGatewayTransportError, not a raw error."""
+
+    def fake_get_server_certificate(addr: object, *, timeout: float | None = None) -> str:
+        raise raised
+
+    monkeypatch.setattr(options_module.ssl, "get_server_certificate", fake_get_server_certificate)
+
+    options = ClientOptions(endpoint="gateway.example:5001")
+    with pytest.raises(MxGatewayTransportError) as excinfo:
+        create_channel(options)
+    assert options.endpoint in str(excinfo.value)

clients/python/tests/test_cli.py

@@ -2,14 +2,79 @@
 
 import json
 
+import pytest
 from click.testing import CliRunner
 
 from zb_mom_ww_mxgateway import __version__
+from zb_mom_ww_mxgateway_cli import commands as commands_module
 from zb_mom_ww_mxgateway_cli.commands import main
 
 _BATCH_EOR = "__MXGW_BATCH_EOR__"
 
 
+def test_require_certificate_validation_flag_flows_through_connect(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    """The --require-certificate-validation CLI flag must reach ClientOptions (Client.Python-027)."""
+    captured: dict[str, object] = {}
+
+    async def fake_connect(options, **_kwargs):
+        captured["options"] = options
+        # Return a minimal object that supports the async context-manager protocol
+        # used by every CLI command body (async with await _connect(...) as client).
+        return _FakeAsyncClient()
+
+    monkeypatch.setattr(commands_module.GatewayClient, "connect", fake_connect)
+
+    result = CliRunner().invoke(
+        main,
+        [
+            "open-session",
+            "--endpoint",
+            "gateway.example:5001",
+            "--require-certificate-validation",
+            "--json",
+        ],
+    )
+
+    assert result.exit_code == 0, result.output
+    assert captured["options"].require_certificate_validation is True
+
+
+def test_require_certificate_validation_defaults_off(monkeypatch: pytest.MonkeyPatch) -> None:
+    """Without the flag the strict-validation posture stays off (TOFU default)."""
+    captured: dict[str, object] = {}
+
+    async def fake_connect(options, **_kwargs):
+        captured["options"] = options
+        return _FakeAsyncClient()
+
+    monkeypatch.setattr(commands_module.GatewayClient, "connect", fake_connect)
+
+    result = CliRunner().invoke(
+        main,
+        ["open-session", "--endpoint", "gateway.example:5001", "--plaintext", "--json"],
+    )
+
+    assert result.exit_code == 0, result.output
+    assert captured["options"].require_certificate_validation is False
+
+
+class _FakeAsyncClient:
+    """Minimal async-context-manager fake satisfying the open-session command body."""
+
+    async def __aenter__(self) -> "_FakeAsyncClient":
+        return self
+
+    async def __aexit__(self, *_exc: object) -> None:
+        return None
+
+    async def open_session_raw(self, *_args, **_kwargs):
+        from zb_mom_ww_mxgateway.generated import mxaccess_gateway_pb2 as pb
+
+        return pb.OpenSessionReply(session_id="cli-test-session")
+
+
 def test_version_json_is_deterministic() -> None:
     runner = CliRunner()
 

clients/python/tests/test_client_session.py

@@ -8,9 +8,107 @@ from typing import Any
 import pytest
 
 from zb_mom_ww_mxgateway import ClientOptions, GatewayClient, MxAccessError
+from zb_mom_ww_mxgateway import client as client_module
+from zb_mom_ww_mxgateway import galaxy as galaxy_module
+from zb_mom_ww_mxgateway.galaxy import GalaxyRepositoryClient
 from zb_mom_ww_mxgateway.generated import mxaccess_gateway_pb2 as pb
 
 
+@pytest.mark.asyncio
+async def test_gateway_connect_forwards_require_certificate_validation(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    """The connect convenience kwarg must reach ClientOptions (Client.Python-027)."""
+    captured: dict[str, Any] = {}
+
+    def fake_create_channel(options: ClientOptions) -> object:
+        captured["options"] = options
+        return object()
+
+    monkeypatch.setattr(client_module, "create_channel", fake_create_channel)
+    monkeypatch.setattr(client_module.pb_grpc, "MxAccessGatewayStub", lambda channel: object())
+
+    await GatewayClient.connect(
+        endpoint="gateway.example:5001",
+        require_certificate_validation=True,
+    )
+
+    assert captured["options"].require_certificate_validation is True
+
+
+@pytest.mark.asyncio
+async def test_galaxy_connect_forwards_require_certificate_validation(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    """GalaxyRepositoryClient.connect must thread the flag too (Client.Python-027)."""
+    captured: dict[str, Any] = {}
+
+    def fake_create_channel(options: ClientOptions) -> object:
+        captured["options"] = options
+        return object()
+
+    monkeypatch.setattr(galaxy_module, "create_channel", fake_create_channel)
+    monkeypatch.setattr(
+        galaxy_module.galaxy_pb_grpc, "GalaxyRepositoryStub", lambda channel: object()
+    )
+
+    await GalaxyRepositoryClient.connect(
+        endpoint="gateway.example:5001",
+        require_certificate_validation=True,
+    )
+
+    assert captured["options"].require_certificate_validation is True
+
+
+@pytest.mark.asyncio
+async def test_gateway_connect_runs_create_channel_off_the_event_loop(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    """connect must run the blocking channel factory off the loop (Client.Python-028)."""
+    ran_in_thread: dict[str, bool] = {}
+
+    def fake_create_channel(options: ClientOptions) -> object:
+        # If this runs on the event loop thread, get_running_loop() succeeds.
+        try:
+            asyncio.get_running_loop()
+            ran_in_thread["off_loop"] = False
+        except RuntimeError:
+            ran_in_thread["off_loop"] = True
+        return object()
+
+    monkeypatch.setattr(client_module, "create_channel", fake_create_channel)
+    monkeypatch.setattr(client_module.pb_grpc, "MxAccessGatewayStub", lambda channel: object())
+
+    await GatewayClient.connect(endpoint="gateway.example:5001")
+
+    assert ran_in_thread["off_loop"] is True
+
+
+@pytest.mark.asyncio
+async def test_galaxy_connect_runs_create_channel_off_the_event_loop(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    """GalaxyRepositoryClient.connect must also run the probe off the loop (Client.Python-028)."""
+    ran_in_thread: dict[str, bool] = {}
+
+    def fake_create_channel(options: ClientOptions) -> object:
+        try:
+            asyncio.get_running_loop()
+            ran_in_thread["off_loop"] = False
+        except RuntimeError:
+            ran_in_thread["off_loop"] = True
+        return object()
+
+    monkeypatch.setattr(galaxy_module, "create_channel", fake_create_channel)
+    monkeypatch.setattr(
+        galaxy_module.galaxy_pb_grpc, "GalaxyRepositoryStub", lambda channel: object()
+    )
+
+    await GalaxyRepositoryClient.connect(endpoint="gateway.example:5001")
+
+    assert ran_in_thread["off_loop"] is True
+
+
 @pytest.mark.asyncio
 async def test_session_helpers_send_auth_metadata_and_preserve_raw_replies() -> None:
     stub = FakeGatewayStub()

clients/python/tests/test_tls.py

@@ -134,6 +134,17 @@ def test_split_authority_parses_host_and_port() -> None:
     assert _split_authority(":5120") == ("localhost", 5120)
 
 
+def test_split_authority_defaults_port_for_portless_endpoint() -> None:
+    from zb_mom_ww_mxgateway.options import _split_authority
+
+    # A bare hostname (no ":port") must default to 443, not crash on int("mygateway").
+    assert _split_authority("mygateway") == ("mygateway", 443)
+    # Scheme-prefixed bare hostname behaves the same.
+    assert _split_authority("https://mygateway") == ("mygateway", 443)
+    # A non-numeric tail after a colon is treated as no explicit port.
+    assert _split_authority("mygateway:") == ("mygateway", 443)
+
+
 def test_split_authority_strips_ipv6_brackets() -> None:
     from zb_mom_ww_mxgateway.options import _split_authority