feat(cli): support conversational AI follow-ups in interactive mode

tests/test_cli.py

@@ -188,7 +188,28 @@ def test_interactive_collect_then_quit(monkeypatch) -> None:  # type: ignore[no-
 def test_interactive_unknown_command_prints_hint(monkeypatch) -> None:  # type: ignore[no-untyped-def]
     _mock_session(monkeypatch)
 
-    commands = iter(["/wat", "/quit"])
+    async def fake_collect_from_plan(_session, _plan) -> CollectionReport:  # type: ignore[no-untyped-def]
+        return CollectionReport(
+            host="ssh.archflux.net",
+            items=[
+                CollectedItem(
+                    name="kernel",
+                    result=SSHCommandResult(
+                        command="uname -a",
+                        exit_code=0,
+                        stdout="Linux test",
+                        stderr="",
+                    ),
+                ),
+            ],
+        )
+
+    commands = iter(["what should I check next?", "/quit"])
+    monkeypatch.setattr("tai.cli.collect_from_plan", fake_collect_from_plan)
+    monkeypatch.setattr(
+        "tai.cli.AIClient.stream_messages",
+        lambda *_args, **_kwargs: iter(["Check logs."]),
+    )
     monkeypatch.setattr("builtins.input", lambda _prompt: next(commands))
 
     runner = CliRunner()
@@ -206,4 +227,5 @@ def test_interactive_unknown_command_prints_hint(monkeypatch) -> None:  # type:
     )
 
     assert result.exit_code == 0
-    assert "Unknown command" in result.stdout
+    assert "Analyzing..." in result.stdout
+    assert "Check logs." in result.stdout