core/cli: update display and state for streaming, error, cancel events

src/plyngent/cli/display.py

@@ -5,6 +5,8 @@ from typing import TYPE_CHECKING
 import click
 
 from plyngent.agent import (
+    CancelledEvent,
+    ErrorEvent,
     MaxRoundsEvent,
     TextDeltaEvent,
     ToolCallEvent,
@@ -20,7 +22,7 @@ if TYPE_CHECKING:
 _TOOL_RESULT_PREVIEW = 200
 
 
-async def render_events(events: AsyncIterator[AgentEvent]) -> None:
+async def render_events(events: AsyncIterator[AgentEvent]) -> None:  # noqa: C901, PLR0912
     """Print agent events to the terminal."""
     printed_text = False
     async for event in events:
@@ -47,6 +49,10 @@ async def render_events(events: AsyncIterator[AgentEvent]) -> None:
                 else content[:_TOOL_RESULT_PREVIEW] + "…"
             )
             click.secho(f"[tool result] {preview}", fg="magenta")
+        elif isinstance(event, ErrorEvent):
+            click.secho(f"\n[error] {event.message}", fg="bright_red")
+        elif isinstance(event, CancelledEvent):
+            click.secho("\n[cancelled]", fg="yellow")
         elif isinstance(event, MaxRoundsEvent):
             if event.continued:
                 click.secho(
@@ -60,5 +66,5 @@ async def render_events(events: AsyncIterator[AgentEvent]) -> None:
             _ = event
     if printed_text:
         click.echo()
-    # Blank line before the next readline prompt so log and input are not jammed.
     click.echo()
+

src/plyngent/cli/state.py

@@ -34,7 +34,7 @@ class ReplState:
 
     def __post_init__(self) -> None:
         # DeepSeek client uses a compatible but distinct param type; treat as ChatClient.
-        self.client = cast("ChatClient", create_client(self.provider))
+        self.client = cast("ChatClient", cast("object", create_client(self.provider)))
         self.agent = self._make_agent()
 
     def _tool_registry(self) -> ToolRegistry | None:
@@ -53,12 +53,13 @@ class ReplState:
             session_id=self.session_id,
             max_rounds=self.max_rounds,
             on_limit=prompt_continue_limit,
+            stream=True,
         )
 
     def rebuild_client(self) -> None:
         """Recreate client and agent after provider/model/tools change."""
         messages = list(self.agent.messages)
-        self.client = cast("ChatClient", create_client(self.provider))
+        self.client = cast("ChatClient", cast("object", create_client(self.provider)))
         self.agent = self._make_agent()
         self.agent.messages = messages
 

tests/test_agent/test_loop.py

@@ -96,10 +96,10 @@ async def test_run_chat_loop_text_only() -> None:
         ]
     )
     messages: list[AnyChatMessage] = [UserChatMessage(content="hi")]
-    events = [e async for e in run_chat_loop(client, messages, model="m")]
+    events = [e async for e in run_chat_loop(client, messages, model="m", stream=False)]
-    assert isinstance(events[0], AssistantMessageEvent)
+    assert isinstance(events[0], TextDeltaEvent)
-    assert isinstance(events[1], TextDeltaEvent)
+    assert events[0].content == "hello"
-    assert events[1].content == "hello"
+    assert isinstance(events[1], AssistantMessageEvent)
     assert len(messages) == 2  # noqa: PLR2004
     assert len(client.calls) == 1