diff --git a/autogpt_platform/backend/backend/copilot/db.py b/autogpt_platform/backend/backend/copilot/db.py
index b6e8e73e29..4e939502e1 100644
--- a/autogpt_platform/backend/backend/copilot/db.py
+++ b/autogpt_platform/backend/backend/copilot/db.py
@@ -163,10 +163,42 @@ async def get_chat_messages_paginated(
                 # very start of the conversation (sequence 0).
                 if boundary_msgs[0].sequence > 0:
                     has_more = True
+    else:
+        # Forward mode: DB returned ASC.
+        # Tool-call tail boundary fix: if the last message in this page is a
+        # tool message, the NEXT forward page would start after it and begin
+        # mid-tool-group — the owning assistant message is on this page but
+        # the following tool results are on the next page.
+        # Trim the current page so it ends on the owning assistant message,
+        # which keeps tool groups intact across page boundaries.
+        if results and results[-1].role == "tool":
+            # Walk backward through results to find the last non-tool message.
+            trim_idx = len(results) - 1
+            while trim_idx >= 0 and results[trim_idx].role == "tool":
+                trim_idx -= 1
+
+            if trim_idx >= 0:
+                # Trim results so the page ends at the owning assistant.
+                # Mark has_more=True so the client knows to fetch the rest.
+                results = results[: trim_idx + 1]
+                has_more = True
+            else:
+                # Entire page is tool messages with no visible owner — log and
+                # keep as-is so the caller is not stuck with an empty page.
+                logger.warning(
+                    "Forward tail boundary: entire page is tool messages "
+                    "for session=%s, no owning assistant found (%d msgs)",
+                    session_id,
+                    len(results),
+                )
 
     messages = [ChatMessage.from_db(m) for m in results]
     oldest_sequence = messages[0].sequence if messages else None
-    newest_sequence = messages[-1].sequence if messages else None
+    # newest_sequence is only meaningful in forward mode; in backward mode it
+    # points to the last message of the page (not the session's newest message)
+    # which is not a valid forward cursor.  Return None in backward mode so
+    # clients don't accidentally use it as one.
+    newest_sequence = messages[-1].sequence if (messages and forward) else None
 
     return PaginatedMessages(
         messages=messages,
diff --git a/autogpt_platform/backend/backend/copilot/db_test.py b/autogpt_platform/backend/backend/copilot/db_test.py
index 8b1dbf43c5..2c7045e751 100644
--- a/autogpt_platform/backend/backend/copilot/db_test.py
+++ b/autogpt_platform/backend/backend/copilot/db_test.py
@@ -271,10 +271,10 @@ async def test_after_sequence_returns_messages_in_order(
 
 
 @pytest.mark.asyncio
-async def test_newest_sequence_populated_for_backward_mode(
+async def test_newest_sequence_none_for_backward_mode(
     mock_db: tuple[AsyncMock, AsyncMock],
 ):
-    """newest_sequence is populated for backward-paginated results."""
+    """newest_sequence is None in backward mode — it is not a valid forward cursor."""
     find_first, _ = mock_db
     find_first.return_value = _make_session(
         messages=[_make_msg(5), _make_msg(4), _make_msg(3)],
@@ -283,7 +283,7 @@ async def test_newest_sequence_populated_for_backward_mode(
     page = await get_chat_messages_paginated(SESSION_ID, limit=50)
 
     assert page is not None
-    assert page.newest_sequence == 5
+    assert page.newest_sequence is None
     assert page.oldest_sequence == 3
 
 
@@ -302,6 +302,56 @@ async def test_forward_mode_no_boundary_expansion(
     assert find_many.call_count == 0
 
 
+@pytest.mark.asyncio
+async def test_forward_tail_boundary_trims_trailing_tool_messages(
+    mock_db: tuple[AsyncMock, AsyncMock],
+):
+    """Forward pages that end with tool messages are trimmed to the owning
+    assistant so the next after_sequence page doesn't start mid-tool-group."""
+    find_first, _ = mock_db
+    # DB returns 4 messages ASC: assistant at 0, tool at 1, tool at 2, tool at 3
+    find_first.return_value = _make_session(
+        messages=[
+            _make_msg(0, role="assistant"),
+            _make_msg(1, role="tool"),
+            _make_msg(2, role="tool"),
+            _make_msg(3, role="tool"),
+        ],
+    )
+
+    page = await get_chat_messages_paginated(SESSION_ID, limit=10, from_start=True)
+
+    assert page is not None
+    # Page should be trimmed to end at the assistant message
+    assert [m.sequence for m in page.messages] == [0]
+    assert page.newest_sequence == 0
+    # has_more must be True so the client fetches the tool messages on next page
+    assert page.has_more is True
+
+
+@pytest.mark.asyncio
+async def test_forward_tail_boundary_no_trim_when_last_not_tool(
+    mock_db: tuple[AsyncMock, AsyncMock],
+):
+    """Forward pages that end with a non-tool message are not trimmed."""
+    find_first, _ = mock_db
+    find_first.return_value = _make_session(
+        messages=[
+            _make_msg(0, role="user"),
+            _make_msg(1, role="assistant"),
+            _make_msg(2, role="tool"),
+            _make_msg(3, role="assistant"),
+        ],
+    )
+
+    page = await get_chat_messages_paginated(SESSION_ID, limit=10, from_start=True)
+
+    assert page is not None
+    assert [m.sequence for m in page.messages] == [0, 1, 2, 3]
+    assert page.newest_sequence == 3
+    assert page.has_more is False
+
+
 @pytest.mark.asyncio
 async def test_user_id_filter_applied_to_session_where(
     mock_db: tuple[AsyncMock, AsyncMock],
diff --git a/autogpt_platform/frontend/src/app/(platform)/copilot/__tests__/useLoadMoreMessages.test.ts b/autogpt_platform/frontend/src/app/(platform)/copilot/__tests__/useLoadMoreMessages.test.ts
index c8d8f0d074..80ed40ece7 100644
--- a/autogpt_platform/frontend/src/app/(platform)/copilot/__tests__/useLoadMoreMessages.test.ts
+++ b/autogpt_platform/frontend/src/app/(platform)/copilot/__tests__/useLoadMoreMessages.test.ts
@@ -358,6 +358,60 @@ describe("useLoadMoreMessages", () => {
 
       expect(result.current.hasMore).toBe(false);
     });
+
+    it("forward truncation keeps first MAX_OLDER_MESSAGES items (not last)", async () => {
+      // 1990 messages already paged; load 20 more forward — total 2010 > 2000.
+      // Forward truncation must keep slice(0, 2000), not slice(-2000),
+      // to preserve the beginning of the conversation.
+      const forwardNearLimitArgs = {
+        ...BASE_ARGS,
+        forwardPaginated: true,
+        initialNewestSequence: 49,
+        initialOldestSequence: 0,
+        initialHasMore: true,
+      };
+
+      const { result } = renderHook((props) => useLoadMoreMessages(props), {
+        initialProps: forwardNearLimitArgs,
+      });
+
+      // First load: 1990 messages — advances newestSequence to 2039
+      mockGetV2GetSession.mockResolvedValueOnce(
+        makeSuccessResponse({
+          messages: Array.from({ length: 1990 }, (_, i) => ({
+            role: "assistant",
+            content: `msg ${i + 50}`,
+            sequence: i + 50,
+          })),
+          has_more_messages: true,
+          newest_sequence: 2039,
+        }),
+      );
+
+      await act(async () => {
+        await result.current.loadMore();
+      });
+
+      // Second load: 20 more messages pushes total to 2010 > 2000 — hasMore=false
+      mockGetV2GetSession.mockResolvedValueOnce(
+        makeSuccessResponse({
+          messages: Array.from({ length: 20 }, (_, i) => ({
+            role: "assistant",
+            content: `msg ${i + 2040}`,
+            sequence: i + 2040,
+          })),
+          has_more_messages: false,
+          newest_sequence: 2059,
+        }),
+      );
+
+      await act(async () => {
+        await result.current.loadMore();
+      });
+
+      // After truncation, hasMore is forced false (total ≥ MAX_OLDER_MESSAGES).
+      expect(result.current.hasMore).toBe(false);
+    });
   });
 
   describe("loadMore — null cursor guard", () => {
diff --git a/autogpt_platform/frontend/src/app/(platform)/copilot/useLoadMoreMessages.ts b/autogpt_platform/frontend/src/app/(platform)/copilot/useLoadMoreMessages.ts
index 340cecfa06..49a13ef8d9 100644
--- a/autogpt_platform/frontend/src/app/(platform)/copilot/useLoadMoreMessages.ts
+++ b/autogpt_platform/frontend/src/app/(platform)/copilot/useLoadMoreMessages.ts
@@ -158,7 +158,15 @@ export function useLoadMoreMessages({
           ? [...prev, ...newRaw]
           : [...newRaw, ...prev];
         if (merged.length > MAX_OLDER_MESSAGES) {
-          return merged.slice(merged.length - MAX_OLDER_MESSAGES);
+          // Backward: discard the oldest (front) items — user has scrolled far
+          // back and we shed the furthest history.
+          // Forward: discard the newest (tail) items — we only ever fetch
+          // forward, so the tail is the most recently appended page; shedding
+          // it means the sentinel stalls, which is safer than discarding the
+          // beginning of the conversation the user is here to read.
+          return forwardPaginated
+            ? merged.slice(0, MAX_OLDER_MESSAGES)
+            : merged.slice(merged.length - MAX_OLDER_MESSAGES);
         }
         return merged;
       });