diff --git a/apps/api/src/services/platform_operator_service.py b/apps/api/src/services/platform_operator_service.py
index 2b80335f..e929bd17 100644
--- a/apps/api/src/services/platform_operator_service.py
+++ b/apps/api/src/services/platform_operator_service.py
@@ -158,7 +158,12 @@ async def list_runs(
         if remediation_status:
             result = await db.execute(stmt)
             candidate_rows = list(result.scalars().all())
-            inbound_by_run, outbound_by_run = await _load_run_message_context(db, candidate_rows)
+            context_limit = _list_filter_context_limit(len(candidate_rows))
+            inbound_by_run, outbound_by_run = await _load_run_message_context(
+                db,
+                candidate_rows,
+                limit=context_limit,
+            )
             remediation_summaries = await _build_run_remediation_summaries(
                 runs=candidate_rows,
                 inbound_by_run=inbound_by_run,
@@ -351,6 +356,8 @@ def _collect_run_incident_ids(
 async def _load_run_message_context(
     db: Any,
     runs: list[AwoooPRunState],
+    *,
+    limit: int = _MAX_LIST_CONTEXT_ROWS,
 ) -> tuple[
     dict[UUID, list[AwoooPConversationEvent]],
     dict[UUID, list[AwoooPOutboundMessage]],
@@ -384,7 +391,7 @@ async def _load_run_message_context(
         select(AwoooPConversationEvent)
         .where(sa_or(*inbound_filters))
         .order_by(AwoooPConversationEvent.received_at.desc())
-        .limit(_MAX_LIST_CONTEXT_ROWS)
+        .limit(limit)
     )
     inbound_by_run: dict[UUID, list[AwoooPConversationEvent]] = defaultdict(list)
     for event in inbound_result.scalars().all():
@@ -400,7 +407,7 @@ async def _load_run_message_context(
         select(AwoooPOutboundMessage)
         .where(AwoooPOutboundMessage.run_id.in_(run_ids))
         .order_by(AwoooPOutboundMessage.queued_at.desc())
-        .limit(_MAX_LIST_CONTEXT_ROWS)
+        .limit(limit)
     )
     outbound_by_run: dict[UUID, list[AwoooPOutboundMessage]] = defaultdict(list)
     for message in outbound_result.scalars().all():
@@ -409,6 +416,10 @@ async def _load_run_message_context(
     return dict(inbound_by_run), dict(outbound_by_run)
 
 
+def _list_filter_context_limit(candidate_count: int) -> int:
+    return min(max(candidate_count * 4, _MAX_LIST_CONTEXT_ROWS), 20_000)
+
+
 def _route_label_from_remediation(item: dict[str, Any]) -> str:
     """Render remediation MCP route consistently with Telegram / Work Items."""
     return "/".join(
diff --git a/apps/api/tests/test_awooop_operator_timeline_labels.py b/apps/api/tests/test_awooop_operator_timeline_labels.py
index d38bbab1..86ffe2ec 100644
--- a/apps/api/tests/test_awooop_operator_timeline_labels.py
+++ b/apps/api/tests/test_awooop_operator_timeline_labels.py
@@ -3,6 +3,7 @@ from types import SimpleNamespace
 
 from src.services.platform_operator_service import (
     _collect_run_incident_ids,
+    _list_filter_context_limit,
     _outbound_timeline_title,
     _run_remediation_list_summary,
     _remediation_summary_matches_status,
@@ -171,6 +172,12 @@ def test_remediation_summary_matches_status_filter() -> None:
     assert _remediation_summary_matches_status(None, "no_evidence")
 
 
+def test_list_filter_context_limit_scales_with_candidate_rows() -> None:
+    assert _list_filter_context_limit(2) == 500
+    assert _list_filter_context_limit(4176) == 16704
+    assert _list_filter_context_limit(10000) == 20000
+
+
 def test_timeline_sort_key_normalizes_datetime_and_iso_string() -> None:
     fallback = datetime(2026, 5, 14, 10, 0, 0)
     keys = [