feat(gateway): surface natural mid-turn assistant messages in chat platforms

Add display.interim_assistant_messages config (enabled by default) that
forwards completed assistant commentary between tool calls to the user
as separate chat messages. Models already emit useful status text like
'I'll inspect the repo first.' — this surfaces it on Telegram, Discord,
and other messaging platforms instead of swallowing it.

Independent from tool_progress and gateway streaming. Disabled for
webhooks. Uses GatewayStreamConsumer when available, falls back to
direct adapter send. Tracks response_previewed to prevent double-delivery
when interim message matches the final response.

Also fixes: cursor not stripped from fallback prefix in stream consumer
(affected continuation calculation on no-edit platforms like Signal).

Cherry-picked from PR #7885 by asheriif, default changed to enabled.
Fixes #5016

tests/run_agent/test_run_agent_codex_responses.py

@@ -744,6 +744,44 @@ def test_normalize_codex_response_marks_commentary_only_message_as_incomplete(mo
     assert "inspect the repository" in (assistant_message.content or "")
 
 
+def test_interim_commentary_is_not_marked_already_streamed_without_callbacks(monkeypatch):
+    agent = _build_agent(monkeypatch)
+    observed = {}
+
+    agent._fire_stream_delta("short version: yes")
+    agent.interim_assistant_callback = lambda text, *, already_streamed=False: observed.update(
+        {"text": text, "already_streamed": already_streamed}
+    )
+
+    agent._emit_interim_assistant_message({"role": "assistant", "content": "short version: yes"})
+
+    assert observed == {
+        "text": "short version: yes",
+        "already_streamed": False,
+    }
+
+
+def test_interim_commentary_is_not_marked_already_streamed_when_stream_callback_fails(monkeypatch):
+    agent = _build_agent(monkeypatch)
+    observed = {}
+
+    def failing_callback(_text):
+        raise RuntimeError("display failed")
+
+    agent.stream_delta_callback = failing_callback
+    agent._fire_stream_delta("short version: yes")
+    agent.interim_assistant_callback = lambda text, *, already_streamed=False: observed.update(
+        {"text": text, "already_streamed": already_streamed}
+    )
+
+    agent._emit_interim_assistant_message({"role": "assistant", "content": "short version: yes"})
+
+    assert observed == {
+        "text": "short version: yes",
+        "already_streamed": False,
+    }
+
+
 def test_run_conversation_codex_continues_after_commentary_phase_message(monkeypatch):
     agent = _build_agent(monkeypatch)
     responses = [