fix(bedrock): don't raise a retryable error on malformed tool call arguments

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

litellm/litellm_core_utils/prompt_templates/factory.py

@@ -3682,7 +3682,10 @@ def _convert_to_bedrock_tool_call_invoke(
                         #   '{"cmd":"a"}{"cmd":"b"}{"cmd":"c"}'
                         # Split them and emit one toolUse block per object.
                         # Fixes: https://github.com/BerriAI/litellm/issues/20543
-                        parsed_objects = split_concatenated_json_objects(arguments)
+                        try:
+                            parsed_objects = split_concatenated_json_objects(arguments)
+                        except json.JSONDecodeError:
+                            parsed_objects = []
                         if parsed_objects:
                             # First object keeps the original tool id.
                             for obj_idx, obj in enumerate(parsed_objects):
@@ -3718,8 +3721,13 @@ def _convert_to_bedrock_tool_call_invoke(
                         _parts_list.append(cache_point_block)
         return _parts_list
     except Exception as e:
-        raise Exception(
-            "Unable to convert openai tool calls={} to bedrock tool calls. Received error={}".format(tool_calls, str(e))
+        tool_call_ids = [tool.get("id") for tool in tool_calls if isinstance(tool, dict)]
+        raise litellm.BadRequestError(
+            message="Unable to convert openai tool calls with ids={} to bedrock tool calls. Received error={}".format(
+                tool_call_ids, str(e)
+            ),
+            model=model or "",
+            llm_provider="bedrock",
         )
 
 

tests/test_litellm/litellm_core_utils/prompt_templates/test_litellm_core_utils_prompt_templates_factory.py

@@ -2286,6 +2286,44 @@ def test_bedrock_tool_call_invoke_non_dict_arguments():
     assert result[0]["toolUse"]["input"] == {}
 
 
+def test_bedrock_tool_call_invoke_truncated_json_arguments():
+    """
+    Truncated tool call arguments (issue #35303) must not raise. A client replaying a
+    partially streamed tool call would otherwise trigger a pre-network exception that the
+    router maps to a retryable APIConnectionError and retries through the fallback graph.
+    """
+    tool_calls = [
+        {
+            "id": "tooluse_MAh2QLVjBRkvi5QJkLQ08V",
+            "type": "function",
+            "function": {
+                "name": "replace_note_content",
+                "arguments": '{"note_id": "999af35c-4061-4ece-8581-7d43fc988ba4", "title": "WG"',
+            },
+        }
+    ]
+    result = _convert_to_bedrock_tool_call_invoke(tool_calls)
+    assert len(result) == 1
+    assert result[0]["toolUse"]["toolUseId"] == "tooluse_MAh2QLVjBRkvi5QJkLQ08V"
+    assert result[0]["toolUse"]["input"] == {}
+
+
+def test_bedrock_tool_call_invoke_unconvertible_raises_non_retryable_bad_request():
+    """
+    Conversion failures are client input errors, so they must surface as a non-retryable
+    BadRequestError instead of a bare Exception that maps to APIConnectionError, and the
+    message must not embed the tool call payload (issue #35303).
+    """
+    tool_calls = [{"id": "call_bad", "type": "function", "function": None}]
+
+    with pytest.raises(litellm.BadRequestError) as exc_info:
+        _convert_to_bedrock_tool_call_invoke(tool_calls)
+
+    assert exc_info.value.status_code == 400
+    assert "call_bad" in str(exc_info.value)
+    assert "function" not in str(exc_info.value).split("Received error=")[0]
+
+
 def test_make_valid_bedrock_tool_name_preserves_hyphens():
     assert make_valid_bedrock_tool_name("my-tool") == "my-tool"
     assert (