fix(proxy_server.py): If master key is set, only master key can be used to generate new keys

litellm/proxy/proxy_server.py

@@ -164,8 +164,14 @@ async def user_api_key_auth(request: Request):
         return 
     try: 
         api_key = await oauth2_scheme(request=request)
+        route = request.url.path
+
         if api_key == master_key: 
             return
+        
+        if route == "/key/generate" and api_key != master_key: 
+            raise Exception(f"If master key is set, only master key can be used to generate new keys")
+
         if api_key in config_cache:
             llm_model_list =  config_cache[api_key].get("model_list", [])
             return
@@ -593,6 +599,11 @@ async def generate_key_fn(request: Request):
             detail={"error": "models param must be a list"},
         )
 
+@router.get("/test")
+async def test_endpoint(request: Request): 
+    return {"route": request.url.path}
+
+#### EXPERIMENTAL QUEUING #### 
 @router.post("/queue/request", dependencies=[Depends(user_api_key_auth)])
 async def async_queue_request(request: Request): 
     global celery_fn, llm_model_list

litellm/tests/test_completion.py

@@ -164,9 +164,7 @@ def test_completion_gpt4_vision():
 
 def test_completion_perplexity_api():
     try:
-        litellm.set_verbose=True
-        litellm.num_retries = 0
-        litellm.drop_params = True
+        # litellm.set_verbose=True
         messages=[{
             "role": "system", 
             "content": "You're a good bot"
@@ -180,7 +178,7 @@ def test_completion_perplexity_api():
         response = completion(
             model="mistral-7b-instruct", 
             messages=messages,
-            api_base="https://api.perplexity.ai", stop="Hello")
+            api_base="https://api.perplexity.ai")
         print(response)
     except Exception as e:
         pytest.fail(f"Error occurred: {e}")
@@ -446,7 +444,7 @@ def test_completion_openai():
         pass
     except Exception as e:
         pytest.fail(f"Error occurred: {e}")
-test_completion_openai()
+# test_completion_openai()
 
 def test_completion_text_openai():
     try:
@@ -832,6 +830,7 @@ def test_completion_replicate_llama2_stream():
             #     assert len(chunk.choices[0].delta["content"]) > 2
             # print(chunk)
         assert len(complete_response) > 5
+        print(f"complete_response: {complete_response}")
     except Exception as e:
         pytest.fail(f"Error occurred: {e}")
 test_completion_replicate_llama2_stream()