test(e2e): address greptile review feedback

Remove the duplicate cache/cache_params block in the gateway config so the two
can't silently diverge under future edits. Reorder the soft-budget test to assert
the call isn't a budget block before require_successful_call, since that helper
hard-fails any non-2xx and left the budget-block check unreachable; the misleading
"skip" comment is corrected. Add a deferred delete in test_budget_delete_removes_it
so a failed delete doesn't leak a budget on the shared proxy. Scope the
spend_tracking sys.path insertion in pytest_sessionfinish to just the cleanup
import so a broader "pytest tests/" run isn't left with a mutated path.

tests/e2e/budgets/test_budget_crud_e2e.py

@@ -38,6 +38,7 @@ def test_budget_crud_roundtrip(client: BudgetClient, resources: ResourceManager)
 
 def test_budget_delete_removes_it(client: BudgetClient, resources: ResourceManager) -> None:
     budget_id = client.create_budget(max_budget=1.0)
+    resources.defer(lambda: client.delete_budget(budget_id))
     client.delete_budget(budget_id)
     assert not client.budget_info(budget_id), "budget still present after delete"
 

tests/e2e/budgets/test_soft_budget_e2e.py

@@ -28,8 +28,8 @@ def test_soft_budget_does_not_block(
         result = client.chat(
             key, "claude-haiku-4-5", f"hi {unique_marker()}", max_tokens=16
         )
-        require_successful_call(result)  # skip if provider unavailable
         assert not is_budget_block(result), (
             "soft_budget blocked a request; it must alert only, not block "
             f"(body={result.body[:200]})"
         )
+        require_successful_call(result)  # any other non-2xx (e.g. provider down) is a hard fail

tests/e2e/conftest.py

@@ -32,14 +32,20 @@ def pytest_configure(config: pytest.Config) -> None:
 def pytest_sessionfinish(session: pytest.Session, exitstatus: int) -> None:
     """Once the whole e2e session is done (all suites), truncate the spend logs so
     the DB doesn't accumulate test rows. Best-effort: a cleanup failure (no DB
-    reachable) must not fail the run."""
-    sys.path.insert(0, str(Path(__file__).parent / "spend_tracking"))
+    reachable) must not fail the run. The spend_tracking dir goes on sys.path only
+    for this import and is removed after, so a broader `pytest tests/` run is not
+    left with a mutated path."""
+    spend_dir = str(Path(__file__).parent / "spend_tracking")
+    sys.path.insert(0, spend_dir)
     try:
         from spend_e2e_client import reset_spend_logs  # pyright: ignore
 
         reset_spend_logs()
     except Exception as exc:  # noqa: BLE001 - cleanup is best-effort
         print(f"spend-log cleanup skipped: {exc}")
+    finally:
+        if spend_dir in sys.path:
+            sys.path.remove(spend_dir)
 
 
 @pytest.fixture(scope="session", autouse=True)

tests/e2e/gateway/litellm-config.yml

@@ -61,14 +61,6 @@ litellm_settings:
   # service_callback: ["datadog"]
   callbacks: ["arize_phoenix", "datadog", "smtp_email", "prometheus", "otel"]
   require_auth_for_metrics_endpoint: false
-  cache: true
-  cache_params:
-    type: redis
-    host: redis
-    port: 6379
-    password: os.environ/REDIS_PASSWORD
-    namespace: litellm.caching
-    ttl: 16600
   #type: redis-semantic
   #similarity_threshold: 0.8   # similarity threshold for semantic cache
   #redis_semantic_cache_embedding_model: text-embedding-ada-002 # only works with text-embedding-ada-002 for now... https://github.com/BerriAI/litellm/issues/4001