diff --git a/providers/common/ai/docs/message_history.rst b/providers/common/ai/docs/message_history.rst
index 453b4d7a7c37e..fa9ee78befee1 100644
--- a/providers/common/ai/docs/message_history.rst
+++ b/providers/common/ai/docs/message_history.rst
@@ -30,6 +30,13 @@ turns and, after the run, pushes the full updated transcript
 run reads it back to resume the conversation. ``None`` (the default) keeps the
 single-turn behavior unchanged.
 
+To record the transcript of a single-turn run, set ``emit_message_history=True``.
+The operator starts the run with no prior turns and pushes ``result.all_messages()``
+to the same ``message_history`` XCom key, so the transcript is available to
+downstream tasks for inspection, evaluation or storage. ``emit_message_history`` defaults
+to ``True`` when ``message_history`` is set; pass ``False`` to keep a seeded run from
+pushing it.
+
 The operator does **not** decide *where* a session is stored -- that keying is
 deployment-specific. The pattern is three tasks: load the prior transcript for
 the session, run the agent, store the updated transcript. The example keys a
@@ -49,7 +56,8 @@ fresh session instead of trying to parse the string ``"None"``.
 
 .. note::
 
-    ``message_history`` cannot be combined with ``enable_hitl_review`` -- the
+    ``message_history`` and ``emit_message_history`` cannot be combined with
+    ``enable_hitl_review`` -- the
     operator raises at construction. The post-review (human-approved) transcript
     is not recoverable today, so emitting the pre-review transcript would
     silently drop the reviewed turns.
diff --git a/providers/common/ai/docs/operators/agent.rst b/providers/common/ai/docs/operators/agent.rst
index fbb7b542989c4..9dbf792c9c9a9 100644
--- a/providers/common/ai/docs/operators/agent.rst
+++ b/providers/common/ai/docs/operators/agent.rst
@@ -451,6 +451,9 @@ Parameters
   When set, the post-run transcript is pushed to XCom under the key
   ``message_history`` for the next run to resume. Default ``None`` (single-turn).
   See :doc:`../message_history`.
+- ``emit_message_history``: Push the post-run transcript to XCom under the key
+  ``message_history``. Defaults to ``True`` when ``message_history`` is set and ``False``
+  otherwise; ``True`` records a single-turn run without seeding it. See :doc:`../message_history`.
 - ``serialize_output``: If ``True`` and ``output_type`` is a Pydantic
   ``BaseModel`` subclass, the model instance is dumped to a ``dict`` via
   ``model_dump()`` before being pushed to XCom. Default ``False`` -- the
diff --git a/providers/common/ai/docs/stability.rst b/providers/common/ai/docs/stability.rst
index 5feb431be94e9..3fad7b4ca4f92 100644
--- a/providers/common/ai/docs/stability.rst
+++ b/providers/common/ai/docs/stability.rst
@@ -69,6 +69,10 @@ Pydantic AI toolsets, but the Pydantic AI class they inherit from can change.
      - Seeds the run with the given conversation, as a list of messages or its JSON form,
        and publishes the finished conversation under the ``message_history`` XCom key so a
        later run can continue it. Cannot be combined with ``enable_hitl_review``.
+   * - ``emit_message_history`` on ``AgentOperator``
+     - Publishes the finished conversation under the ``message_history`` XCom key without
+       seeding the run. Defaults to ``True`` when ``message_history`` is set, and ``False``
+       stops a seeded run from publishing it. Cannot be combined with ``enable_hitl_review``.
    * - ``durable=True`` on ``AgentOperator`` and ``@task.agent``, and
        :class:`~airflow.providers.common.ai.durable.AirflowDurability`
        (:doc:`durable_execution`)
diff --git a/providers/common/ai/docs/troubleshooting.rst b/providers/common/ai/docs/troubleshooting.rst
index ea3e8c3fae59e..e2b329e74dcfa 100644
--- a/providers/common/ai/docs/troubleshooting.rst
+++ b/providers/common/ai/docs/troubleshooting.rst
@@ -130,7 +130,7 @@ as a task failure.
     ``AgentOperator(durable=True)`` adds the capability to the agent, so a second one in
     ``capabilities`` would record every step twice. Drop one of them.
 
-``message_history and enable_hitl_review=True cannot be used together``
+``message_history and enable_hitl_review=True cannot be used together`` and ``emit_message_history and enable_hitl_review=True cannot be used together``
     The post-review transcript is not recoverable today, so the operator refuses rather
     than silently dropping the reviewed turns. See :doc:`message_history`.
 
diff --git a/providers/common/ai/src/airflow/providers/common/ai/operators/agent.py b/providers/common/ai/src/airflow/providers/common/ai/operators/agent.py
index 06f052cb6c295..2392a1fdd19c0 100644
--- a/providers/common/ai/src/airflow/providers/common/ai/operators/agent.py
+++ b/providers/common/ai/src/airflow/providers/common/ai/operators/agent.py
@@ -406,6 +406,11 @@ class AgentOperator(CancellableAgentRunMixin, BaseOperator, HITLReviewMixin, Usa
         object-storage XCom backend or trim old turns. Not supported together with
         ``enable_hitl_review`` (raises) -- the post-review transcript is not yet
         recoverable.
+    :param emit_message_history: Push the post-run transcript to XCom under the key
+        ``message_history``. Defaults to ``True`` when ``message_history`` is set and ``False``
+        otherwise. Set ``True`` to record a single-turn run without seeding it, or ``False``
+        to keep a seeded run from pushing it. Not supported together with
+        ``enable_hitl_review`` (raises).
 
     **HITL Review parameters** (requires the ``hitl_review`` plugin):
 
@@ -507,6 +512,7 @@ def __init__(
         durable: bool = False,
         cache_prompt: bool = True,
         message_history: list[ModelMessage] | str | bytes | None = None,
+        emit_message_history: bool | None = None,
         # Agent feedback parameters
         enable_hitl_review: bool = False,
         max_hitl_iterations: int = 5,
@@ -537,6 +543,9 @@ def __init__(
         # No validation here -- see coerce_usage_limits() docstring for why.
         self.usage_limits = usage_limits
         self.message_history = message_history
+        self.emit_message_history = (
+            message_history is not None if emit_message_history is None else emit_message_history
+        )
 
         self.durable = durable
         self.cache_prompt = cache_prompt
@@ -581,6 +590,9 @@ def __init__(
             # surface the final message history.
             raise ValueError("message_history and enable_hitl_review=True cannot be used together.")
 
+        if emit_message_history and enable_hitl_review:
+            raise ValueError("emit_message_history and enable_hitl_review=True cannot be used together.")
+
         if durable or enable_hitl_review:
             self._reject_sandbox_without_continuity(durable=durable, enable_hitl_review=enable_hitl_review)
 
@@ -1078,7 +1090,7 @@ def _complete_run(self, context: Context, result: Any, *, attempt_usage: RunUsag
         if self._usage_budget and (run_usage := self._run_usage) is not None:
             self._log_cumulative_usage(run_usage)
 
-        if self.message_history is not None:
+        if self.emit_message_history:
             self._emit_message_history(context, result)
 
         output = result.output
diff --git a/providers/common/ai/tests/unit/common/ai/operators/test_agent.py b/providers/common/ai/tests/unit/common/ai/operators/test_agent.py
index 70d7faa9e4121..178db39816679 100644
--- a/providers/common/ai/tests/unit/common/ai/operators/test_agent.py
+++ b/providers/common/ai/tests/unit/common/ai/operators/test_agent.py
@@ -2127,22 +2127,32 @@ def test_none_is_single_turn_no_history_no_emit(self, mock_hook_cls, make_mock_r
         assert "message_history" not in pushed_keys
         assert pushed_keys == {"run_id", "usage", MODEL_NAME_XCOM_KEY}
 
+    @pytest.mark.parametrize(
+        ("kwargs", "emits"),
+        [
+            pytest.param({"message_history": []}, True, id="seeded"),
+            pytest.param({"emit_message_history": True}, True, id="opt-in"),
+            pytest.param({"message_history": [], "emit_message_history": False}, False, id="opt-out"),
+        ],
+    )
     @patch("airflow.providers.common.ai.operators.agent.PydanticAIHook", autospec=True)
-    def test_transcript_emitted_to_xcom_when_history_set(self, mock_hook_cls, make_mock_run_result):
-        """When message_history is set, the post-run transcript is pushed to XCom and round-trips."""
+    def test_transcript_emitted_to_xcom(self, mock_hook_cls, kwargs, emits, make_mock_run_result):
+        """The transcript push follows emit_message_history, which defaults to whether the run is seeded."""
         mock_agent = _make_mock_agent("ok", make_mock_run_result)
         mock_agent.run_sync.return_value.all_messages.return_value = _sample_history()
         mock_hook_cls.get_hook.return_value.create_agent.return_value = mock_agent
 
-        op = AgentOperator(task_id="t", prompt="run", llm_conn_id="c", message_history=[])
+        op = AgentOperator(task_id="t", prompt="run", llm_conn_id="c", **kwargs)
         context = _make_context()
         op.execute(context=context)
 
+        # emit_message_history never seeds the run.
+        assert ("message_history" in mock_agent.run_sync.call_args.kwargs) is ("message_history" in kwargs)
         ti = context["task_instance"]
         pushes = {c.kwargs["key"]: c.kwargs["value"] for c in ti.xcom_push.call_args_list}
-        assert set(pushes) == {"run_id", "usage", "message_history", MODEL_NAME_XCOM_KEY}
-        restored = ModelMessagesTypeAdapter.validate_json(pushes["message_history"])
-        assert len(restored) == 2
+        assert ("message_history" in pushes) is emits
+        if emits:
+            assert len(ModelMessagesTypeAdapter.validate_json(pushes["message_history"])) == 2
 
     @patch("airflow.providers.common.ai.operators.agent.PydanticAIHook", autospec=True)
     def test_usage_limits_still_forwarded_with_history(
@@ -2165,16 +2175,19 @@ def test_usage_limits_still_forwarded_with_history(
     @pytest.mark.skipif(
         not AIRFLOW_V_3_1_PLUS, reason="Human in the loop is only compatible with Airflow >= 3.1.0"
     )
-    def test_message_history_with_hitl_review_raises(self):
-        """message_history cannot be combined with HITL review (post-review transcript is lost)."""
-        with pytest.raises(ValueError, match="message_history and enable_hitl_review"):
-            AgentOperator(
-                task_id="t",
-                prompt="run",
-                llm_conn_id="c",
-                message_history=[],
-                enable_hitl_review=True,
-            )
+    @pytest.mark.parametrize(
+        ("kwargs", "match"),
+        [
+            pytest.param({"message_history": []}, "message_history and enable_hitl_review", id="seeded"),
+            pytest.param(
+                {"emit_message_history": True}, "emit_message_history and enable_hitl_review", id="opt-in"
+            ),
+        ],
+    )
+    def test_message_history_with_hitl_review_raises(self, kwargs, match):
+        """Transcript emission cannot be combined with HITL review (post-review transcript is lost)."""
+        with pytest.raises(ValueError, match=match):
+            AgentOperator(task_id="t", prompt="run", llm_conn_id="c", enable_hitl_review=True, **kwargs)
 
     def test_durable_path_also_seeds_message_history(self):
         """The durable branch forwards message_history into the run, and keeps its conversation id."""