From 0691ff8d136b445eb316be2f7b134d1bafa70c8a Mon Sep 17 00:00:00 2001
From: "Jugal D. Bhatt" <55304795+jugaldb@users.noreply.github.com>
Date: Fri, 11 Jul 2025 04:24:48 +0530
Subject: [PATCH] Added validate payload error (#12494)

* Added validate payload error

* added logger

* added test case
---
 .../proxy/common_utils/http_parsing_utils.py  | 24 +++--
 .../common_utils/test_http_parsing_utils.py   | 87 +++++++++++++++++++
 2 files changed, 105 insertions(+), 6 deletions(-)

diff --git a/litellm/proxy/common_utils/http_parsing_utils.py b/litellm/proxy/common_utils/http_parsing_utils.py
index 6ded2f000c..875d69437c 100644
--- a/litellm/proxy/common_utils/http_parsing_utils.py
+++ b/litellm/proxy/common_utils/http_parsing_utils.py
@@ -5,6 +5,7 @@ import orjson
 from fastapi import Request, UploadFile, status
 
 from litellm._logging import verbose_proxy_logger
+from litellm.proxy._types import ProxyException
 from litellm.types.router import Deployment
 
 
@@ -44,8 +45,8 @@ async def _read_request_body(request: Optional[Request]) -> Dict:
             else:
                 try:
                     parsed_body = orjson.loads(body)
-                except orjson.JSONDecodeError:
-                    # Fall back to the standard json module which is more forgiving
+                except orjson.JSONDecodeError as e:
+                    # First try the standard json module which is more forgiving
                     # First decode bytes to string if needed
                     body_str = body.decode("utf-8") if isinstance(body, bytes) else body
 
@@ -61,15 +62,26 @@ async def _read_request_body(request: Optional[Request]) -> Dict:
                         r"(?<![\uD800-\uDBFF])[\uDC00-\uDFFF]", "", body_str
                     )
 
-                    parsed_body = json.loads(body_str)
+                    try:
+                        parsed_body = json.loads(body_str)
+                    except json.JSONDecodeError:
+                        # If both orjson and json.loads fail, throw a proper error
+                        verbose_proxy_logger.error(f"Invalid JSON payload received: {str(e)}")
+                        raise ProxyException(
+                            message=f"Invalid JSON payload: {str(e)}",
+                            type="invalid_request_error",
+                            param="request_body",
+                            code=status.HTTP_400_BAD_REQUEST,
+                        )
 
         # Cache the parsed result
         _safe_set_request_parsed_body(request=request, parsed_body=parsed_body)
         return parsed_body
 
-    except (json.JSONDecodeError, orjson.JSONDecodeError):
-        verbose_proxy_logger.exception("Invalid JSON payload received.")
-        return {}
+    except (json.JSONDecodeError, orjson.JSONDecodeError, ProxyException) as e:
+        # Re-raise ProxyException as-is
+        verbose_proxy_logger.error(f"Invalid JSON payload received: {str(e)}")
+        raise
     except Exception as e:
         # Catch unexpected errors to avoid crashes
         verbose_proxy_logger.exception(
diff --git a/tests/test_litellm/proxy/common_utils/test_http_parsing_utils.py b/tests/test_litellm/proxy/common_utils/test_http_parsing_utils.py
index b98a1f2657..721bcfc377 100644
--- a/tests/test_litellm/proxy/common_utils/test_http_parsing_utils.py
+++ b/tests/test_litellm/proxy/common_utils/test_http_parsing_utils.py
@@ -20,6 +20,7 @@ from litellm.proxy.common_utils.http_parsing_utils import (
     _safe_set_request_parsed_body,
     get_form_data,
 )
+from litellm.proxy._types import ProxyException
 
 
 @pytest.mark.asyncio
@@ -150,6 +151,92 @@ async def test_circular_reference_handling():
     )  # This will pass, showing the cache pollution
 
 
+@pytest.mark.asyncio
+async def test_json_parsing_error_handling():
+    """
+    Test that JSON parsing errors are properly handled and raise ProxyException
+    with appropriate error messages.
+    """
+    # Test case 1: Trailing comma error
+    mock_request = MagicMock()
+    invalid_json_with_trailing_comma = b'''{
+        "model": "gpt-4o",
+        "tools": [
+            {
+                "type": "mcp",
+                "server_label": "litellm",
+                "headers": {
+                    "x-litellm-api-key": "Bearer sk-1234",
+                }
+            }
+        ],
+        "input": "Run available tools"
+    }'''
+    
+    mock_request.body = AsyncMock(return_value=invalid_json_with_trailing_comma)
+    mock_request.headers = {"content-type": "application/json"}
+    mock_request.scope = {}
+
+    # Should raise ProxyException for trailing comma
+    with pytest.raises(ProxyException) as exc_info:
+        await _read_request_body(mock_request)
+    
+    assert exc_info.value.code == "400"
+    assert "Invalid JSON payload" in exc_info.value.message
+    assert "trailing comma" in exc_info.value.message
+
+    # Test case 2: Unquoted property name error
+    mock_request2 = MagicMock()
+    invalid_json_unquoted_property = b'''{
+        "model": "gpt-4o",
+        "tools": [
+            {
+                type: "mcp",
+                "server_label": "litellm"
+            }
+        ],
+        "input": "Run available tools"
+    }'''
+    
+    mock_request2.body = AsyncMock(return_value=invalid_json_unquoted_property)
+    mock_request2.headers = {"content-type": "application/json"}
+    mock_request2.scope = {}
+
+    # Should raise ProxyException for unquoted property
+    with pytest.raises(ProxyException) as exc_info2:
+        await _read_request_body(mock_request2)
+    
+    assert exc_info2.value.code == "400"
+    assert "Invalid JSON payload" in exc_info2.value.message
+
+    # Test case 3: Valid JSON should work normally
+    mock_request3 = MagicMock()
+    valid_json = b'''{
+        "model": "gpt-4o",
+        "tools": [
+            {
+                "type": "mcp",
+                "server_label": "litellm",
+                "headers": {
+                    "x-litellm-api-key": "Bearer sk-1234"
+                }
+            }
+        ],
+        "input": "Run available tools"
+    }'''
+    
+    mock_request3.body = AsyncMock(return_value=valid_json)
+    mock_request3.headers = {"content-type": "application/json"}
+    mock_request3.scope = {}
+
+    # Should parse successfully
+    result = await _read_request_body(mock_request3)
+    assert result["model"] == "gpt-4o"
+    assert result["input"] == "Run available tools"
+    assert len(result["tools"]) == 1
+    assert result["tools"][0]["type"] == "mcp"
+
+
 @pytest.mark.asyncio
 async def test_get_form_data():
     """