LCORE-1253: Address radofuchs review feedback

Rewrite proxy e2e tests to follow project conventions:

- Reuse existing step definitions: use "I use query to ask question"
  from llm_query_response.py and "The status code of the response is"
  from common_http.py instead of custom query/response steps.
- Split service restart into two explicit Given steps: "Llama Stack
  is restarted" and "Lightspeed Stack is restarted" so restart
  ordering is visible in the feature file.
- Remove local (non-Docker) mode code path. Proxy tests use
  restart_container() exclusively, consistent with the rest of the
  e2e test suite.
- Check specific status code 500 for error scenarios instead of the
  broad >= 400 range.
- Remove custom send_query, verify_llm_response, and
  verify_error_response steps that duplicated existing functionality.

Net reduction: -183 lines from step definitions.

Author: max-svistunov

tests/e2e/features/proxy.feature

@@ -17,9 +17,13 @@ Feature: Proxy and TLS networking tests for Llama Stack providers
   Scenario: LLM traffic is routed through a configured tunnel proxy
     Given A tunnel proxy is running on port 8888
       And Llama Stack is configured to route inference through the tunnel proxy
-      And The services are restarted with the modified Llama Stack config
-     When I send a query "What is 2+2?" to the LLM
-     Then The LLM responds successfully
+      And Llama Stack is restarted
+      And Lightspeed Stack is restarted
+     When I use "query" to ask question
+    """
+    {"query": "What is 2+2?", "model": "{MODEL}", "provider": "{PROVIDER}"}
+    """
+     Then The status code of the response is 200
       And The tunnel proxy handled at least 1 CONNECT request to the LLM provider
 
   # NOTE: no_proxy is defined on Llama Stack's ProxyConfig model but not
@@ -29,9 +33,13 @@ Feature: Proxy and TLS networking tests for Llama Stack providers
   @TunnelProxy
   Scenario: LLM query fails gracefully when proxy is unreachable
     Given Llama Stack is configured to route inference through proxy "http://127.0.0.1:19999"
-      And The services are restarted with the modified Llama Stack config
-     When I send a query "What is 2+2?" to the LLM
-     Then The response indicates a proxy or connection error
+      And Llama Stack is restarted
+      And Lightspeed Stack is restarted
+     When I use "query" to ask question
+    """
+    {"query": "What is 2+2?", "model": "{MODEL}", "provider": "{PROVIDER}"}
+    """
+     Then The status code of the response is 500
 
 
   # --- AC2: Interception proxy with CA certificate ---
@@ -40,39 +48,59 @@ Feature: Proxy and TLS networking tests for Llama Stack providers
   Scenario: LLM traffic works through interception proxy with correct CA
     Given An interception proxy with trustme CA is running on port 8889
       And Llama Stack is configured to route inference through the interception proxy with CA cert
-      And The services are restarted with the modified Llama Stack config
-     When I send a query "What is 2+2?" to the LLM
-     Then The LLM responds successfully
+      And Llama Stack is restarted
+      And Lightspeed Stack is restarted
+     When I use "query" to ask question
+    """
+    {"query": "What is 2+2?", "model": "{MODEL}", "provider": "{PROVIDER}"}
+    """
+     Then The status code of the response is 200
       And The interception proxy intercepted at least 1 connection
 
   @InterceptionProxy
   Scenario: LLM query fails when interception proxy CA is not provided
     Given An interception proxy with trustme CA is running on port 8890
       And Llama Stack is configured to route inference through the interception proxy without CA cert
-      And The services are restarted with the modified Llama Stack config
-     When I send a query "What is 2+2?" to the LLM
-     Then The response indicates a proxy or connection error
+      And Llama Stack is restarted
+      And Lightspeed Stack is restarted
+     When I use "query" to ask question
+    """
+    {"query": "What is 2+2?", "model": "{MODEL}", "provider": "{PROVIDER}"}
+    """
+     Then The status code of the response is 500
 
 
   # --- AC3: TLS version and cipher configuration ---
 
   @TLSVersion
   Scenario: TLS minimum version TLSv1.2 is respected
     Given Llama Stack is configured with minimum TLS version "TLSv1.2"
-      And The services are restarted with the modified Llama Stack config
-     When I send a query "What is 2+2?" to the LLM
-     Then The LLM responds successfully
+      And Llama Stack is restarted
+      And Lightspeed Stack is restarted
+     When I use "query" to ask question
+    """
+    {"query": "What is 2+2?", "model": "{MODEL}", "provider": "{PROVIDER}"}
+    """
+     Then The status code of the response is 200
 
   @TLSVersion
   Scenario: TLS minimum version TLSv1.3 is respected
     Given Llama Stack is configured with minimum TLS version "TLSv1.3"
-      And The services are restarted with the modified Llama Stack config
-     When I send a query "What is 2+2?" to the LLM
-     Then The LLM responds successfully
+      And Llama Stack is restarted
+      And Lightspeed Stack is restarted
+     When I use "query" to ask question
+    """
+    {"query": "What is 2+2?", "model": "{MODEL}", "provider": "{PROVIDER}"}
+    """
+     Then The status code of the response is 200
 
   @TLSCipher
   Scenario: Custom cipher suite configuration is respected
     Given Llama Stack is configured with ciphers "ECDHE+AESGCM:DHE+AESGCM"
-      And The services are restarted with the modified Llama Stack config
-     When I send a query "What is 2+2?" to the LLM
-     Then The LLM responds successfully
+      And Llama Stack is restarted
+      And Lightspeed Stack is restarted
+     When I use "query" to ask question
+    """
+    {"query": "What is 2+2?", "model": "{MODEL}", "provider": "{PROVIDER}"}
+    """
+     Then The status code of the response is 200