feat: rewrite controlplane integration tests with pytest

.github/workflows/integration-testing.yml

@@ -119,7 +119,7 @@ jobs:
             ignore: ""
           # TODO: expand to full tests_integ/memory once test stability is addressed
           - group: memory
-            path: tests_integ/memory/test_memory_client.py
+            path: tests_integ/memory/test_controlplane.py tests_integ/memory/test_memory_client.py
             timeout: 10
             extra-deps: ""
             ignore: ""
@@ -151,7 +151,7 @@ jobs:
       - name: Install dependencies
         run: |
           pip install -e .
-          pip install --no-cache-dir pytest requests strands-agents uvicorn httpx starlette websockets ${{ matrix.extra-deps }}
+          pip install --no-cache-dir pytest pytest-xdist pytest-order requests strands-agents uvicorn httpx starlette websockets ${{ matrix.extra-deps }}
 
       - name: Run integration tests
         env:
@@ -162,7 +162,7 @@ jobs:
         id: tests
         timeout-minutes: ${{ matrix.timeout }}
         run: |
-          pytest ${{ matrix.path }} ${{ matrix.ignore }} -s --log-cli-level=INFO
+          pytest ${{ matrix.path }} ${{ matrix.ignore }} -n auto --dist=loadscope -s --log-cli-level=INFO
 
   safety-gate:
     runs-on: ubuntu-latest

pyproject.toml

@@ -144,6 +144,7 @@ dev = [
     "pytest>=8.4.1",
     "pytest-asyncio>=0.24.0",
     "pytest-cov>=6.0.0",
+    "pytest-order>=1.3.0",
     "ruff>=0.12.0",
     "websockets>=14.1",
     "wheel>=0.45.1",

tests_integ/memory/helpers.py

@@ -0,0 +1,16 @@
+import time
+
+import pytest
+
+
+def poll_until(fn, predicate, max_wait=180, poll_interval=10, fail_if=None):
+    """Poll fn() until predicate(result) is True or max_wait exceeded."""
+    deadline = time.time() + max_wait
+    while time.time() < deadline:
+        result = fn()
+        if fail_if and fail_if(result):
+            pytest.fail(f"poll_until hit terminal state: {result}")
+        if predicate(result):
+            return result
+        time.sleep(poll_interval)
+    pytest.fail(f"poll_until timed out after {max_wait}s")