fix(compression): anchor merged-replay detection on the last summary marker

_has_merged_inflight_replay partitioned on the FIRST end marker. A
merged-into-tail carrier can embed an older carrier (marker + replay
header + task) in its prior-context block, ahead of the new summary's own
marker, so a carrier with nothing after its real boundary was misread as
already carrying the active request. _find_inflight_user_task would stop
there and the reappend would drag the new summary body into the task
text. rpartition anchors on the real boundary; summary bodies never keep
an inner marker and the _force_user_leading layout has exactly one, so
every legitimate layout still matches.

Also compute the stripped remainder once and use removeprefix instead of
a manual length slice.

Co-authored-by: ppazosp <pablopazosp3@gmail.com>

agent/context_compressor.py

@@ -4762,11 +4762,15 @@ Write only the summary body. Do not include any preamble or prefix."""
         if not cls._is_context_summary_message(message):
             return False
         text = _content_text_for_contains(message.get("content"))
-        _, boundary, remainder = text.partition(_SUMMARY_END_MARKER)
+        # The LAST end marker is the real handoff boundary: a merged-into-tail
+        # carrier can embed an older carrier (marker + replay) in its prior
+        # context, ahead of the new summary's own marker.
+        _, boundary, remainder = text.rpartition(_SUMMARY_END_MARKER)
+        rest = remainder.lstrip()
         return bool(
             boundary
-            and remainder.lstrip().startswith(_INFLIGHT_TASK_REPLAY_HEADER)
-            and remainder.lstrip()[len(_INFLIGHT_TASK_REPLAY_HEADER):].strip()
+            and rest.startswith(_INFLIGHT_TASK_REPLAY_HEADER)
+            and rest.removeprefix(_INFLIGHT_TASK_REPLAY_HEADER).strip()
         )
 
     @classmethod