From dc3941e331eed6d4f32e3a74cecc84f43038bbdd Mon Sep 17 00:00:00 2001
From: Utsab Dahal <134686066+utsab345@users.noreply.github.com>
Date: Thu, 2 Oct 2025 07:30:58 +0545
Subject: [PATCH] fix: prevent JSON content from being added multiple times
 during retries (#1255)

---
 rdagent/core/experiment.py  |  9 ++++-----
 rdagent/oai/backend/base.py |  5 ++++-
 test/oai/test_base.py       | 23 +++++++++++++++++++++++
 3 files changed, 31 insertions(+), 6 deletions(-)
 create mode 100644 test/oai/test_base.py

diff --git a/rdagent/core/experiment.py b/rdagent/core/experiment.py
index a84159be..f9dac480 100644
--- a/rdagent/core/experiment.py
+++ b/rdagent/core/experiment.py
@@ -13,7 +13,7 @@ from collections.abc import Sequence
 from copy import deepcopy
 from dataclasses import dataclass
 from pathlib import Path
-from typing import TYPE_CHECKING, Any, Generic, List, TypeVar
+from typing import TYPE_CHECKING, Any, Generic, TypeVar
 
 from rdagent.core.conf import RD_AGENT_SETTINGS
 from rdagent.core.evaluation import Feedback
@@ -48,12 +48,11 @@ class AbsTask(ABC):
         """
 
 
-class UserInstructions(List[str]):
+class UserInstructions(list[str]):
     def __str__(self) -> str:
         if self:
             return ("\nUser Instructions (Top priority!):\n" + "\n".join(f"- {ui}" for ui in self)) if self else ""
-        else:
-            return ""
+        return ""
 
 
 class Task(AbsTask):
@@ -69,7 +68,7 @@ class Task(AbsTask):
         self.user_instructions = user_instructions
 
     def get_task_information(self) -> str:
-        return f"Task Name: {self.name}\nDescription: {self.description}{str(self.user_instructions)}"
+        return f"Task Name: {self.name}\nDescription: {self.description}{self.user_instructions!s}"
 
     def __repr__(self) -> str:
         return f"<{self.__class__.__name__} {self.name}>"
diff --git a/rdagent/oai/backend/base.py b/rdagent/oai/backend/base.py
index cf0512a1..79664b9c 100644
--- a/rdagent/oai/backend/base.py
+++ b/rdagent/oai/backend/base.py
@@ -597,9 +597,12 @@ class APIBackend(ABC):
         new_messages = deepcopy(messages)
         # Loop to get a full response
         try_n = 6
+        # Before retry loop, initialize the flag
+        json_added = False
         for _ in range(try_n):  # for some long code, 3 times may not enough for reasoning models
-            if response_format == {"type": "json_object"} and add_json_in_prompt:
+            if response_format == {"type": "json_object"} and add_json_in_prompt and not json_added:
                 self._add_json_in_prompt(new_messages)
+                json_added = True
             response, finish_reason = self._create_chat_completion_inner_function(
                 messages=new_messages,
                 response_format=response_format,
diff --git a/test/oai/test_base.py b/test/oai/test_base.py
new file mode 100644
index 00000000..eb04b2b1
--- /dev/null
+++ b/test/oai/test_base.py
@@ -0,0 +1,23 @@
+import pytest
+
+
+class MockBackend:
+    def __init__(self):
+        self.messages = []
+
+    def _add_json_in_prompt(self, new_messages):
+        self.messages.append("JSON_ADDED")
+
+
+def test_json_added_once():
+    backend = MockBackend()
+    try_n = 3
+    json_added = False
+    new_messages = ["msg1"]
+
+    for _ in range(try_n):
+        if not json_added:
+            backend._add_json_in_prompt(new_messages)
+            json_added = True
+
+    assert backend.messages.count("JSON_ADDED") == 1