diff --git a/tests/test_sprint29.py b/tests/test_sprint29.py
index 4050666e..9a6e0da1 100644
--- a/tests/test_sprint29.py
+++ b/tests/test_sprint29.py
@@ -712,7 +712,15 @@ class TestSSRFCheck:
 
 class TestENVLock:
     def test_env_lock_importable_from_streaming(self):
-        """_ENV_LOCK must be an importable threading-style reentrant lock."""
+        """_ENV_LOCK must be an importable threading-style lock.
+
+        Stage-360 maintainer fix: _ENV_LOCK MUST be non-reentrant
+        (threading.Lock, not RLock) — see QA test_env_lock_is_non_reentrant.
+        RLock would mask the deadlock pattern that the lock is designed to
+        catch. Background workers that need profile env should use the
+        narrow-lock pattern in profile_env_for_background_worker() rather
+        than relying on reentrance.
+        """
         from api.streaming import _ENV_LOCK
 
         assert hasattr(_ENV_LOCK, "acquire"), "_ENV_LOCK must expose acquire()"
@@ -720,10 +728,15 @@ class TestENVLock:
         assert hasattr(_ENV_LOCK, "__enter__"), "_ENV_LOCK must support context manager use"
         assert hasattr(_ENV_LOCK, "__exit__"), "_ENV_LOCK must support context manager use"
 
+        # Verify non-reentrance: a same-thread second acquire(blocking=False)
+        # while the lock is held must fail. This invariant matters because
+        # it catches a class of deadlock bugs early.
         with _ENV_LOCK:
-            acquired = _ENV_LOCK.acquire(False)
-            assert acquired, "_ENV_LOCK must allow reentrant acquisition"
-            _ENV_LOCK.release()
+            acquired_again = _ENV_LOCK.acquire(False)
+            assert not acquired_again, (
+                "_ENV_LOCK must be non-reentrant (threading.Lock, not RLock). "
+                "See QA test_env_lock_is_non_reentrant for the architectural reason."
+            )
 
     def test_env_lock_importable_in_routes(self):
         """api.routes must be able to import _ENV_LOCK from api.streaming."""