fix(restart): deliver completion after channel reconnects (#4931)

docs/channel-plugin-guide.md

@@ -197,7 +197,19 @@ The agent receives the message and processes it. Replies arrive in your `send()`
 |--------|-------------|
 | `async start()` | **Must block forever.** Connect to platform, listen for messages, call `_handle_message()` on each. If this returns, the channel is dead. |
 | `async stop()` | Set `self._running = False` and clean up. Called when gateway shuts down. |
-| `async send(msg: OutboundMessage)` | Deliver an outbound message to the platform. |
+| `async send(msg: OutboundMessage)` | Deliver an outbound message to the platform. Raise when the transport does not accept it. |
+
+#### Outbound delivery contract
+
+A normal return from `send()` means either the visible payload was accepted by the
+platform transport/API, or the channel deliberately had nothing to deliver (for example,
+an empty progress event). Do not log and return when the client is disconnected, still
+starting, or the platform rejects the request. Raise an exception so `ChannelManager` can
+apply the shared retry policy.
+
+`send()` may run as soon as `is_running` becomes true. If a channel sets `_running` before
+its transport is ready, it must keep raising until delivery can be attempted safely. Small
+platform-specific retries are fine, but the final failure must still reach the manager.
 
 ### Interactive Login
 

nanobot/channels/dingtalk.py

@@ -710,10 +710,11 @@ class DingTalkChannel(BaseChannel):
         """Send a message through DingTalk."""
         token = await self._get_access_token()
         if not token:
-            return
+            raise RuntimeError("DingTalk access token unavailable")
 
         if msg.content and msg.content.strip():
-            await self._send_markdown_text(token, msg.chat_id, msg.content.strip())
+            if not await self._send_markdown_text(token, msg.chat_id, msg.content.strip()):
+                raise RuntimeError("DingTalk text message was not delivered")
 
         for media_ref in msg.media or []:
             ok = await self._send_media_ref(token, msg.chat_id, media_ref)
@@ -722,11 +723,12 @@ class DingTalkChannel(BaseChannel):
             self.logger.error("media send failed for {}", media_ref)
             # Send visible fallback so failures are observable by the user.
             filename = self._guess_filename(media_ref, self._guess_upload_type(media_ref))
-            await self._send_markdown_text(
+            if not await self._send_markdown_text(
                 token,
                 msg.chat_id,
                 f"[Attachment send failed: {filename}]",
-            )
+            ):
+                raise RuntimeError("DingTalk attachment fallback was not delivered")
 
     async def _on_message(
         self,

nanobot/channels/discord.py

@@ -264,7 +264,7 @@ if DISCORD_AVAILABLE:
                     channel = await self.fetch_channel(channel_id)
                 except Exception as e:
                     self._channel.logger.warning("channel {} unavailable: {}", msg.chat_id, e)
-                    return
+                    raise
 
             reference, mention_settings = self._build_reply_context(channel, msg.reply_to)
             sent_media = False
@@ -466,8 +466,7 @@ class DiscordChannel(BaseChannel):
         """Send a message through Discord using discord.py."""
         client = self._client
         if client is None or not client.is_ready():
-            self.logger.warning("client not ready; dropping outbound message")
-            return
+            raise RuntimeError("Discord client is not ready")
 
         is_progress = isinstance(msg.event, ProgressEvent)
 

nanobot/channels/feishu.py

@@ -2407,7 +2407,14 @@ class FeishuChannel(BaseChannel):
                         if ok:
                             return
                     # Fall back to regular send if reply fails
-                self._send_message_sync(receive_id_type, msg.chat_id, m_type, content)
+                message_id = self._send_message_sync(
+                    receive_id_type,
+                    msg.chat_id,
+                    m_type,
+                    content,
+                )
+                if not message_id:
+                    raise RuntimeError(f"Feishu {m_type} message was not delivered")
 
             for file_path in msg.media:
                 if not os.path.isfile(file_path):

nanobot/channels/manager.py

@@ -28,7 +28,11 @@ from nanobot.channels._feishu_instances import ChannelInstanceSpec, feishu_insta
 from nanobot.channels.base import BaseChannel
 from nanobot.channels.registry import DEFAULT_ENABLED_CHANNELS
 from nanobot.config.schema import Config
-from nanobot.utils.restart import consume_restart_notice_from_env, format_restart_completed_message
+from nanobot.utils.restart import (
+    RestartNotice,
+    consume_restart_notice_from_env,
+    format_restart_completed_message,
+)
 
 if TYPE_CHECKING:
     from nanobot.session.manager import SessionManager
@@ -46,6 +50,8 @@ def _default_webui_dist() -> Path | None:
 
 # Retry delays for message sending (exponential backoff: 1s, 2s, 4s)
 _SEND_RETRY_DELAYS = (1, 2, 4)
+_RESTART_NOTICE_START_TIMEOUT_S = 30.0
+_RESTART_NOTICE_START_POLL_S = 0.25
 
 _BOOL_CAMEL_ALIASES: dict[str, str] = {
     "send_progress": "sendProgress",
@@ -476,15 +482,40 @@ class ChannelManager:
         # Wait for all to complete (they should run forever)
         await asyncio.gather(*tasks, return_exceptions=True)
 
-    def _notify_restart_done_if_needed(self) -> None:
-        """Send restart completion message when runtime env markers are present."""
+    def _notify_restart_done_if_needed(self) -> asyncio.Task[None] | None:
+        """Schedule restart completion after the target channel starts."""
         notice = consume_restart_notice_from_env()
         if not notice:
-            return
+            return None
+        return asyncio.create_task(self._send_restart_notice_when_started(notice))
+
+    async def _send_restart_notice_when_started(
+        self,
+        notice: RestartNotice,
+        *,
+        timeout_s: float = _RESTART_NOTICE_START_TIMEOUT_S,
+        poll_s: float = _RESTART_NOTICE_START_POLL_S,
+    ) -> None:
+        """Deliver a restart notice after the target channel starts."""
+        loop = asyncio.get_running_loop()
+        deadline = loop.time() + timeout_s
         target = self.channels.get(notice.channel)
-        if not target:
+        if target is None:
+            logger.warning("Restart notice target channel is not enabled: {}", notice.channel)
             return
-        asyncio.create_task(self._send_with_retry(
+
+        while not target.is_running:
+            remaining = deadline - loop.time()
+            if remaining <= 0:
+                logger.warning(
+                    "Restart notice target did not start: {}:{}",
+                    notice.channel,
+                    notice.chat_id,
+                )
+                return
+            await asyncio.sleep(min(poll_s, remaining))
+
+        await self._send_with_retry(
             target,
             OutboundMessage(
                 channel=notice.channel,
@@ -492,7 +523,8 @@ class ChannelManager:
                 content=format_restart_completed_message(notice.started_at_raw),
                 metadata=dict(notice.metadata or {}),
             ),
-        ))
+            deadline=deadline,
+        )
 
     async def stop_all(self) -> None:
         """Stop all channels and the dispatcher."""
@@ -791,30 +823,52 @@ class ChannelManager:
         merged = replace_outbound_event(first_msg, final_event, content=combined_content)
         return merged, non_matching
 
-    async def _send_with_retry(self, channel: BaseChannel, msg: OutboundMessage) -> None:
+    async def _send_with_retry(
+        self,
+        channel: BaseChannel,
+        msg: OutboundMessage,
+        *,
+        deadline: float | None = None,
+    ) -> None:
         """Send a message with retry on failure using exponential backoff.
 
+        When deadline is provided, retry until that monotonic time instead of
+        stopping at the configured attempt limit.
+
         Note: CancelledError is re-raised to allow graceful shutdown.
         """
         max_attempts = max(self.config.channels.send_max_retries, 1)
+        attempt = 0
 
-        for attempt in range(max_attempts):
+        while True:
+            attempt += 1
             try:
                 await self._send_once(channel, msg)
                 return  # Send succeeded
             except asyncio.CancelledError:
                 raise  # Propagate cancellation for graceful shutdown
             except Exception as e:
-                if attempt == max_attempts - 1:
+                loop = asyncio.get_running_loop()
+                exhausted = (
+                    attempt >= max_attempts
+                    if deadline is None
+                    else loop.time() >= deadline
+                )
+                if exhausted:
                     logger.exception(
                         "Failed to send to {} after {} attempts",
-                        msg.channel, max_attempts
+                        msg.channel, attempt,
                     )
                     return
-                delay = _SEND_RETRY_DELAYS[min(attempt, len(_SEND_RETRY_DELAYS) - 1)]
+                delay = _SEND_RETRY_DELAYS[min(attempt - 1, len(_SEND_RETRY_DELAYS) - 1)]
+                if deadline is not None:
+                    delay = min(delay, max(0.0, deadline - loop.time()))
+                attempt_label = str(attempt)
+                if deadline is None:
+                    attempt_label = f"{attempt}/{max_attempts}"
                 logger.warning(
-                    "Send to {} failed (attempt {}/{}): {}, retrying in {}s",
-                    msg.channel, attempt + 1, max_attempts, type(e).__name__, delay
+                    "Send to {} failed (attempt {}): {}, retrying in {}s",
+                    msg.channel, attempt_label, type(e).__name__, delay,
                 )
                 try:
                     await asyncio.sleep(delay)

nanobot/channels/matrix.py

@@ -559,7 +559,7 @@ class MatrixChannel(BaseChannel):
     async def send(self, msg: OutboundMessage) -> None:
         """Send outbound content; clear typing for non-progress messages."""
         if not self.client:
-            return
+            raise RuntimeError("Matrix client not initialized")
         text = msg.content or ""
         candidates = self._collect_outbound_media_candidates(msg.media)
         relates_to = self._build_thread_relates_to(msg.metadata)
@@ -582,7 +582,9 @@ class MatrixChannel(BaseChannel):
                 content = _build_matrix_text_content(text)
                 if relates_to:
                     content["m.relates_to"] = relates_to
-                await self._send_room_content(msg.chat_id, content)
+                response = await self._send_room_content(msg.chat_id, content)
+                if isinstance(response, RoomSendError):
+                    raise RuntimeError(f"Matrix message was not delivered: {response}")
         finally:
             if not is_progress:
                 await self._stop_typing_keepalive(msg.chat_id, clear_typing=True)

nanobot/channels/napcat.py

@@ -430,8 +430,7 @@ class NapcatChannel(BaseChannel):
 
     async def send(self, msg: OutboundMessage) -> None:
         if self._ws is None:
-            logger.warning("napcat: not connected, dropping outbound message")
-            return
+            raise RuntimeError("napcat: not connected")
 
         kind, _, target = msg.chat_id.partition(":")
         if kind not in ("private", "group") or not target:

nanobot/channels/qq.py

@@ -243,8 +243,7 @@ class QQChannel(BaseChannel):
         """Send attachments first, then text."""
         try:
             if not self._client:
-                self.logger.warning("client not initialized")
-                return
+                raise RuntimeError("QQ client not initialized")
 
             msg_id = msg.metadata.get("message_id")
             chat_type = self._chat_type_cache.get(msg.chat_id, "c2c")
@@ -284,6 +283,7 @@ class QQChannel(BaseChannel):
             raise
         except Exception:
             self.logger.exception("Error sending message to chat_id={}", msg.chat_id)
+            raise
 
     async def _send_text_only(
         self,

nanobot/channels/wecom.py

@@ -493,8 +493,7 @@ class WecomChannel(BaseChannel):
     async def send(self, msg: OutboundMessage) -> None:
         """Send a message through WeCom."""
         if not self._client:
-            self.logger.warning("client not initialized")
-            return
+            raise RuntimeError("WeCom client not initialized")
 
         try:
             content = (msg.content or "").strip()
@@ -553,3 +552,4 @@ class WecomChannel(BaseChannel):
 
         except Exception:
             self.logger.exception("Error sending message to chat_id={}", msg.chat_id)
+            raise

nanobot/utils/restart.py

@@ -9,6 +9,8 @@ from contextlib import suppress
 from dataclasses import dataclass, field
 from typing import Any
 
+from nanobot.webui.metadata import WEBUI_TURN_METADATA_KEY
+
 RESTART_NOTIFY_CHANNEL_ENV = "NANOBOT_RESTART_NOTIFY_CHANNEL"
 RESTART_NOTIFY_CHAT_ID_ENV = "NANOBOT_RESTART_NOTIFY_CHAT_ID"
 RESTART_NOTIFY_METADATA_ENV = "NANOBOT_RESTART_NOTIFY_METADATA"
@@ -40,9 +42,14 @@ def set_restart_notice_to_env(
     os.environ[RESTART_NOTIFY_CHANNEL_ENV] = channel
     os.environ[RESTART_NOTIFY_CHAT_ID_ENV] = chat_id
     os.environ[RESTART_STARTED_AT_ENV] = str(time.time())
-    if metadata:
+    persisted_metadata = dict(metadata or {})
+    persisted_metadata.pop(WEBUI_TURN_METADATA_KEY, None)
+    if persisted_metadata:
         try:
-            os.environ[RESTART_NOTIFY_METADATA_ENV] = json.dumps(metadata, default=str)
+            os.environ[RESTART_NOTIFY_METADATA_ENV] = json.dumps(
+                persisted_metadata,
+                default=str,
+            )
         except (TypeError, ValueError):
             os.environ.pop(RESTART_NOTIFY_METADATA_ENV, None)
     else:

tests/channels/test_channel_plugins.py

@@ -2651,8 +2651,8 @@ async def test_start_all_creates_dispatch_task():
 
 
 @pytest.mark.asyncio
-async def test_notify_restart_done_enqueues_outbound_message():
-    """Restart notice should schedule send_with_retry for target channel."""
+async def test_notify_restart_done_waits_until_channel_starts():
+    """Restart notice should not be sent before the target channel starts."""
     fake_config = SimpleNamespace(
         channels=ChannelsConfig(),
         providers=SimpleNamespace(groq=SimpleNamespace(api_key="")),
@@ -2661,18 +2661,61 @@ async def test_notify_restart_done_enqueues_outbound_message():
     mgr = ChannelManager.__new__(ChannelManager)
     mgr.config = fake_config
     mgr.bus = MessageBus()
-    mgr.channels = {"feishu": _StartableChannel(fake_config, mgr.bus)}
+    channel = _StartableChannel(fake_config, mgr.bus)
+    mgr.channels = {"feishu": channel}
     mgr._dispatch_task = None
     mgr._send_with_retry = AsyncMock()
 
     notice = RestartNotice(channel="feishu", chat_id="oc_123", started_at_raw="100.0")
     with patch("nanobot.channels.manager.consume_restart_notice_from_env", return_value=notice):
-        mgr._notify_restart_done_if_needed()
+        task = mgr._notify_restart_done_if_needed()
 
     await asyncio.sleep(0)
+    mgr._send_with_retry.assert_not_awaited()
+
+    channel._running = True
+    assert task is not None
+    await asyncio.wait_for(task, timeout=1.0)
+
     mgr._send_with_retry.assert_awaited_once()
     sent_channel, sent_msg = mgr._send_with_retry.await_args.args
-    assert sent_channel is mgr.channels["feishu"]
+    assert sent_channel is channel
     assert sent_msg.channel == "feishu"
     assert sent_msg.chat_id == "oc_123"
     assert sent_msg.content.startswith("Restart completed")
+
+
+@pytest.mark.asyncio
+async def test_restart_notice_retries_until_running_channel_accepts_delivery():
+    """A running flag must not make an early transport failure final."""
+
+    class _EventuallyDeliverableChannel(_StartableChannel):
+        def __init__(self, config, bus):
+            super().__init__(config, bus)
+            self.attempts = 0
+            self.sent: OutboundMessage | None = None
+
+        async def send(self, msg: OutboundMessage) -> None:
+            self.attempts += 1
+            if self.attempts == 1:
+                raise RuntimeError("transport not ready")
+            self.sent = msg
+
+    fake_config = SimpleNamespace(
+        channels=ChannelsConfig(send_max_retries=1),
+        providers=SimpleNamespace(groq=SimpleNamespace(api_key="")),
+    )
+    mgr = ChannelManager.__new__(ChannelManager)
+    mgr.config = fake_config
+    mgr.bus = MessageBus()
+    channel = _EventuallyDeliverableChannel(fake_config, mgr.bus)
+    channel._running = True
+    mgr.channels = {"discord": channel}
+
+    notice = RestartNotice(channel="discord", chat_id="123", started_at_raw="")
+    with patch("nanobot.channels.manager._SEND_RETRY_DELAYS", (0,)):
+        await mgr._send_restart_notice_when_started(notice, timeout_s=0.1, poll_s=0.01)
+
+    assert channel.attempts == 2
+    assert channel.sent is not None
+    assert channel.sent.content == "Restart completed."

tests/channels/test_dingtalk_channel.py

@@ -2,6 +2,7 @@ import asyncio
 import zipfile
 from io import BytesIO
 from types import SimpleNamespace
+from unittest.mock import AsyncMock
 
 import httpx
 import pytest
@@ -17,6 +18,7 @@ if not DINGTALK_AVAILABLE:
     pytest.skip("DingTalk dependencies not installed (dingtalk-stream)", allow_module_level=True)
 
 import nanobot.channels.dingtalk as dingtalk_module
+from nanobot.bus.events import OutboundMessage
 from nanobot.bus.queue import MessageBus
 from nanobot.channels.dingtalk import DingTalkChannel, DingTalkConfig, NanobotDingTalkHandler
 
@@ -864,6 +866,35 @@ async def test_send_batch_message_returns_false_on_api_error() -> None:
     assert result is True
 
 
+@pytest.mark.asyncio
+async def test_send_raises_when_access_token_is_unavailable(monkeypatch) -> None:
+    channel = DingTalkChannel(
+        DingTalkConfig(client_id="app", client_secret="secret", allow_from=["*"]),
+        MessageBus(),
+    )
+    monkeypatch.setattr(channel, "_get_access_token", AsyncMock(return_value=None))
+
+    with pytest.raises(RuntimeError, match="access token unavailable"):
+        await channel.send(
+            OutboundMessage(channel="dingtalk", chat_id="user123", content="hello")
+        )
+
+
+@pytest.mark.asyncio
+async def test_send_raises_when_text_is_not_delivered(monkeypatch) -> None:
+    channel = DingTalkChannel(
+        DingTalkConfig(client_id="app", client_secret="secret", allow_from=["*"]),
+        MessageBus(),
+    )
+    monkeypatch.setattr(channel, "_get_access_token", AsyncMock(return_value="token"))
+    monkeypatch.setattr(channel, "_send_markdown_text", AsyncMock(return_value=False))
+
+    with pytest.raises(RuntimeError, match="text message was not delivered"):
+        await channel.send(
+            OutboundMessage(channel="dingtalk", chat_id="user123", content="hello")
+        )
+
+
 @pytest.mark.asyncio
 async def test_send_media_ref_short_circuits_on_transport_error() -> None:
     """When the first send fails with a transport error, _send_media_ref must

tests/channels/test_discord_channel.py

@@ -659,18 +659,19 @@ async def test_on_message_marks_failed_attachment_download(tmp_path, monkeypatch
 
 
 @pytest.mark.asyncio
-async def test_send_warns_when_client_not_ready() -> None:
-    # Sending without a running/ready client should be a safe no-op.
+async def test_send_raises_when_client_not_ready() -> None:
+    # The manager must be able to retry while Discord is still connecting.
     channel = DiscordChannel(DiscordConfig(enabled=True, allow_from=["*"]), MessageBus())
 
-    await channel.send(OutboundMessage(channel="discord", chat_id="123", content="hello"))
+    with pytest.raises(RuntimeError, match="client is not ready"):
+        await channel.send(OutboundMessage(channel="discord", chat_id="123", content="hello"))
 
     assert channel._typing_tasks == {}
 
 
 @pytest.mark.asyncio
-async def test_send_skips_when_channel_not_cached() -> None:
-    # Outbound sends should be skipped when the destination channel is not resolvable.
+async def test_send_raises_when_channel_cannot_be_resolved() -> None:
+    # The manager must be able to retry transient channel-resolution failures.
     owner = DiscordChannel(DiscordConfig(enabled=True, allow_from=["*"]), MessageBus())
     client = DiscordBotClient(owner, intents=discord.Intents.none())
     fetch_calls: list[int] = []
@@ -681,7 +682,10 @@ async def test_send_skips_when_channel_not_cached() -> None:
 
     client.fetch_channel = fetch_channel  # type: ignore[method-assign]
 
-    await client.send_outbound(OutboundMessage(channel="discord", chat_id="123", content="hello"))
+    with pytest.raises(RuntimeError, match="not found"):
+        await client.send_outbound(
+            OutboundMessage(channel="discord", chat_id="123", content="hello")
+        )
 
     assert client.get_channel(123) is None
     assert fetch_calls == [123]

tests/channels/test_feishu_reply.py

@@ -266,8 +266,9 @@ async def test_send_uses_expected_feishu_msg_type_for_uploaded_files(
 
     send_calls: list[tuple[str, str, str, str]] = []
 
-    def _record_send(receive_id_type: str, receive_id: str, msg_type: str, content: str) -> None:
+    def _record_send(receive_id_type: str, receive_id: str, msg_type: str, content: str) -> str:
         send_calls.append((receive_id_type, receive_id, msg_type, content))
+        return "om_test"
 
     with patch.object(channel, "_upload_file_sync", return_value="file-key"), patch.object(
         channel, "_send_message_sync", side_effect=_record_send
@@ -398,6 +399,22 @@ async def test_send_fallback_to_create_when_reply_fails() -> None:
     channel._client.im.v1.message.create.assert_called_once()
 
 
+@pytest.mark.asyncio
+async def test_send_raises_when_create_api_does_not_deliver() -> None:
+    channel = _make_feishu_channel()
+
+    with patch.object(channel, "_send_message_sync", return_value=None):
+        with pytest.raises(RuntimeError, match="message was not delivered"):
+            await channel.send(
+                OutboundMessage(
+                    channel="feishu",
+                    chat_id="oc_abc",
+                    content="hello",
+                    metadata={},
+                )
+            )
+
+
 def test_send_message_sync_falls_back_to_text_for_interactive_error() -> None:
     channel = _make_feishu_channel()
 

tests/channels/test_matrix_channel.py

@@ -1820,6 +1820,28 @@ async def test_send_room_content_returns_room_send_response():
     assert result is client.room_send_response
 
 
+@pytest.mark.asyncio
+async def test_send_raises_when_room_send_returns_error(monkeypatch) -> None:
+    class _FakeRoomSendError:
+        def __str__(self) -> str:
+            return "temporary homeserver failure"
+
+    client = _FakeAsyncClient("", "", "", None)
+    client.room_send_response = _FakeRoomSendError()
+    channel = MatrixChannel(_make_config(), MessageBus())
+    channel.client = client
+    monkeypatch.setattr(matrix_module, "RoomSendError", _FakeRoomSendError)
+
+    with pytest.raises(RuntimeError, match="temporary homeserver failure"):
+        await channel.send(
+            OutboundMessage(
+                channel="matrix",
+                chat_id="!room:matrix.org",
+                content="hello",
+            )
+        )
+
+
 @pytest.mark.asyncio
 async def test_send_delta_creates_stream_buffer_and_sends_initial_message() -> None:
     channel = MatrixChannel(_make_config(), MessageBus())

tests/channels/test_napcat_channel.py

@@ -2,6 +2,7 @@ import asyncio
 
 import pytest
 
+from nanobot.bus.events import OutboundMessage
 from nanobot.bus.queue import MessageBus
 from nanobot.channels.napcat import NapcatChannel, NapcatConfig
 
@@ -52,6 +53,16 @@ def _channel(config: NapcatConfig | None = None) -> NapcatChannel:
     return NapcatChannel(config or NapcatConfig(allow_from=["*"]), MessageBus())
 
 
+@pytest.mark.asyncio
+async def test_send_raises_while_websocket_is_not_connected() -> None:
+    channel = _channel()
+
+    with pytest.raises(RuntimeError, match="not connected"):
+        await channel.send(
+            OutboundMessage(channel="napcat", chat_id="private:123", content="hello")
+        )
+
+
 @pytest.mark.asyncio
 async def test_group_message_requires_mention_by_default() -> None:
     channel = _channel(NapcatConfig(allow_from=["user1"], group_policy="mention"))

tests/channels/test_qq_media.py

@@ -113,17 +113,18 @@ def test_guess_send_file_type_by_mime() -> None:
 
 
 @pytest.mark.asyncio
-async def test_send_exception_caught_not_raised() -> None:
-    """Exceptions inside send() must not propagate."""
+async def test_send_exception_propagates_for_manager_retry() -> None:
+    """Delivery failures must propagate to the channel manager."""
     channel = QQChannel(QQConfig(app_id="app", secret="secret", allow_from=["*"]), MessageBus())
     channel._client = _FakeClient()
 
     with patch.object(
         channel, "_send_text_only", new_callable=AsyncMock, side_effect=RuntimeError("boom")
     ) as send_text:
-        await channel.send(
-            OutboundMessage(channel="qq", chat_id="user1", content="hello")
-        )
+        with pytest.raises(RuntimeError, match="boom"):
+            await channel.send(
+                OutboundMessage(channel="qq", chat_id="user1", content="hello")
+            )
     send_text.assert_awaited_once()
 
 

tests/channels/test_wecom_channel.py

@@ -412,8 +412,8 @@ async def test_send_media_file_not_found() -> None:
 
 
 @pytest.mark.asyncio
-async def test_send_exception_caught_not_raised() -> None:
-    """Exceptions inside send() must not propagate."""
+async def test_send_exception_propagates_for_manager_retry() -> None:
+    """Delivery failures must propagate to the channel manager."""
     channel = WecomChannel(WecomConfig(bot_id="b", secret="s", allow_from=["*"]), MessageBus())
     client = _FakeWeComClient()
     channel._client = client
@@ -423,9 +423,10 @@ async def test_send_exception_caught_not_raised() -> None:
     # Make reply_stream raise
     client.reply_stream.side_effect = RuntimeError("boom")
 
-    await channel.send(
-        OutboundMessage(channel="wecom", chat_id="chat1", content="fail test")
-    )
+    with pytest.raises(RuntimeError, match="boom"):
+        await channel.send(
+            OutboundMessage(channel="wecom", chat_id="chat1", content="fail test")
+        )
     client.reply_stream.assert_called_once()
 
 

tests/utils/test_restart.py

@@ -56,6 +56,23 @@ def test_restart_notice_preserves_metadata_across_env(monkeypatch):
     assert "NANOBOT_RESTART_NOTIFY_METADATA" not in os.environ
 
 
+def test_restart_notice_drops_process_local_webui_turn_metadata(monkeypatch):
+    monkeypatch.delenv("NANOBOT_RESTART_NOTIFY_METADATA", raising=False)
+
+    set_restart_notice_to_env(
+        channel="websocket",
+        chat_id="chat-1",
+        metadata={
+            "webui_turn_id": "turn-from-old-process",
+            "slack": {"thread_ts": "1700.42"},
+        },
+    )
+
+    notice = consume_restart_notice_from_env()
+    assert notice is not None
+    assert notice.metadata == {"slack": {"thread_ts": "1700.42"}}
+
+
 def test_restart_notice_clears_stale_metadata(monkeypatch):
     monkeypatch.setenv("NANOBOT_RESTART_NOTIFY_METADATA", '{"stale": true}')
     set_restart_notice_to_env(channel="cli", chat_id="direct")