Prevent direct OpenAI bypass during inference

.env.example

@@ -1,6 +1,7 @@
-API_BASE_URL=https://api.openai.com/v1
+API_BASE_URL=
 MODEL_NAME=gpt-4o-mini
 API_KEY=
 HF_TOKEN=
 LOCAL_IMAGE_NAME=
 ENV_BASE_URL=http://127.0.0.1:8000
+ALLOW_DIRECT_OPENAI=0

inference.py

@@ -44,6 +44,7 @@ HF_TOKEN = os.getenv("HF_TOKEN")
 PROXY_API_KEY = API_KEY or HF_TOKEN
 LOCAL_IMAGE_NAME = os.getenv("LOCAL_IMAGE_NAME")
 ENV_BASE_URL = os.getenv("ENV_BASE_URL")
+ALLOW_DIRECT_OPENAI = os.getenv("ALLOW_DIRECT_OPENAI") == "1"
 
 BENCHMARK = "support_queue_env"
 SUCCESS_SCORE_THRESHOLD = 0.80
@@ -75,6 +76,13 @@ def create_openai_client() -> Any:
     if not PROXY_API_KEY:
         return None
 
+    if "api.openai.com" in API_BASE_URL and not ALLOW_DIRECT_OPENAI:
+        print(
+            "[DEBUG] Refusing to use direct OpenAI base URL. Set API_BASE_URL to the provided proxy, or set ALLOW_DIRECT_OPENAI=1 for local-only testing.",
+            flush=True,
+        )
+        return None
+
     if OpenAI is not None:
         return OpenAI(base_url=API_BASE_URL, api_key=PROXY_API_KEY)