Merge branch 'feat/optimal-test-suite' into feat/v0.4

.github/workflows/python-tests.yaml

@@ -1,6 +1,19 @@
 # :: Use nektos/act to run this locally
 # :: Example: 
-# :: `act push -j python-tests --matrix python-version:3.10 --container-architecture linux/amd64`
+# :: `act push -j unit-tests --matrix python-version:3.10 --container-architecture linux/amd64`
+#
+# This workflow runs two separate test suites:
+# 1. Unit Tests (python-tests job):
+#    - Runs across Python 3.9 to 3.13
+#    - Located in tests/unit directory
+#    - Coverage report uploaded to Codecov for Python 3.11 only
+#
+# 2. Integration Tests (integration-tests job):
+#    - Runs only on Python 3.13
+#    - Located in tests/integration directory
+#    - Longer timeout (15 min vs 10 min for unit tests)
+#    - Separate cache for dependencies
+
 name: Python Tests
 on:
   workflow_dispatch: {}
@@ -23,10 +36,12 @@ on:
       - 'tests/**/*.ipynb'
 
 jobs:
-  python-tests:
+  unit-tests:
     runs-on: ubuntu-latest
     env:
       OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+      AGENTOPS_API_KEY: ${{ secrets.AGENTOPS_API_KEY }}
+      PYTHONUNBUFFERED: "1"
 
     strategy:
       matrix:
@@ -49,14 +64,10 @@ jobs:
         run: |
           uv sync --group test --group dev
 
-      - name: Run tests with coverage
-        timeout-minutes: 10
+      - name: Run unit tests with coverage
+        timeout-minutes: 5
         run: |
-          uv run -m pytest tests/ -v --cov=agentops --cov-report=xml
-        env:
-          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          AGENTOPS_API_KEY: ${{ secrets.AGENTOPS_API_KEY }}
-          PYTHONUNBUFFERED: "1"
+          uv run -m pytest tests/unit -v --cov=agentops --cov-report=xml
 
       # Only upload coverage report for python3.11
       - name: Upload coverage to Codecov
@@ -68,3 +79,31 @@ jobs:
           flags: unittests
           name: codecov-umbrella
           fail_ci_if_error: true # Should we?
+
+  integration-tests:
+    runs-on: ubuntu-latest
+    env:
+      OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+      AGENTOPS_API_KEY: ${{ secrets.AGENTOPS_API_KEY }}
+      PYTHONUNBUFFERED: "1"
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Setup UV
+        uses: astral-sh/setup-uv@v5
+        continue-on-error: true
+        with:
+          python-version: "3.13"
+          enable-cache: true
+          cache-suffix: uv-3.13-integration
+          cache-dependency-glob: "**/pyproject.toml"
+
+      - name: Install dependencies
+        run: |
+          uv sync --group test --group dev
+
+      - name: Run integration tests
+        timeout-minutes: 5
+        run: |
+          uv run pytest tests/integration

pyproject.toml

@@ -44,27 +44,28 @@ test = [
     "openai>=1.0.0,<2.0.0",
     "langchain",
     "pytest-cov",
+    "fastapi[standard]",
 ]
 
 dev = [
     # Testing essentials
-    "pytest>=7.4.0,<8.0.0",          # Testing framework with good async support
-    "pytest-depends",         # For testing complex agent workflows
-    "pytest-asyncio",         # Async test support for testing concurrent agent operations
-    "pytest-mock",           # Mocking capabilities for isolating agent components
-    "pyfakefs",             # File system testing
-    "pytest-recording",      # Alternative to pytest-vcr with better Python 3.x support
-    "vcrpy @ git+https://github.com/kevin1024/vcrpy.git@81978659f1b18bbb7040ceb324a19114e4a4f328",
+    "pytest>=7.4.0,<8.0.0", # Testing framework with good async support
+    "pytest-depends", # For testing complex agent workflows
+    "pytest-asyncio", # Async test support for testing concurrent agent operations
+    "pytest-mock", # Mocking capabilities for isolating agent components
+    "pyfakefs", # File system testing
+    "pytest-recording", # Alternative to pytest-vcr with better Python 3.x support
+    "vcrpy @ git+https://github.com/kevin1024/vcrpy.git@5f1b20c4ca4a18c1fc8cfe049d7df12ca0659c9b",
     # Code quality and type checking
-    "ruff",                  # Fast Python linter for maintaining code quality
-    "mypy",                # Static type checking for better reliability
-    "types-requests",      # Type stubs for requests library
-
+    "ruff", # Fast Python linter for maintaining code quality
+    "mypy", # Static type checking for better reliability
+    "types-requests", # Type stubs for requests library
     # HTTP mocking and environment
     "requests_mock>=1.11.0", # Mock HTTP requests for testing agent external communications
-    "python-dotenv",         # Environment management for secure testing
-
+    "python-dotenv", # Environment management for secure testing
     # Agent integration testing
+    "pytest-sugar>=1.0.0",
+    "pdbpp>=0.10.3",
 ]
 
 # CI dependencies
@@ -97,11 +98,9 @@ max_line_length = 120
 
 [tool.pytest.ini_options]
 asyncio_mode = "auto"
-asyncio_default_fixture_loop_scope = "function" # WARNING: Changing this may break tests. A `module`-scoped session might be faster, but also unstable.
-test_paths = [
-    "tests",
-]
-addopts = "--tb=short -p no:warnings"
+asyncio_default_fixture_loop_scope = "module" # WARNING: Changing this may break tests. A `module`-scoped session might be faster, but also unstable.
+testpaths = ["tests/unit"]  # Default to unit tests
+addopts = "--tb=short -p no:warnings --import-mode=importlib --ignore=tests/integration"  # Ignore integration by default
 pythonpath = ["."]
 faulthandler_timeout = 30  # Reduced from 60
 timeout = 60  # Reduced from 300

tests/fixtures/event.py

@@ -0,0 +1,32 @@
+from collections import defaultdict
+from typing import TYPE_CHECKING
+
+import pytest
+
+if TYPE_CHECKING:
+    from pytest_mock import MockerFixture
+
+
+@pytest.fixture(scope="function")
+def llm_event_spy(agentops_client, mocker: "MockerFixture") -> dict[str, "MockerFixture"]:
+    """
+    Fixture that provides spies on both providers' response handling
+
+    These fixtures are reset on each test run (function scope). To use it,
+    simply pass it as an argument to the test function. Example:
+
+    ```
+    def test_my_test(llm_event_spy):
+        # test code here
+        llm_event_spy["litellm"].assert_called_once()
+    ```
+    """
+    from agentops.llms.providers.anthropic import AnthropicProvider
+    from agentops.llms.providers.litellm import LiteLLMProvider
+    from agentops.llms.providers.openai import OpenAiProvider
+
+    return {
+        "litellm": mocker.spy(LiteLLMProvider(agentops_client), "handle_response"),
+        "openai": mocker.spy(OpenAiProvider(agentops_client), "handle_response"),
+        "anthropic": mocker.spy(AnthropicProvider(agentops_client), "handle_response"),
+    }

tests/fixtures/packaging.py

@@ -0,0 +1,26 @@
+import builtins
+import pytest
+
+
+@pytest.fixture
+def hide_available_pkg(monkeypatch):
+    """
+    Hide the availability of a package by mocking the __import__ function.
+
+    Usage:
+        @pytest.mark.usefixtures('hide_available_pkg')
+        def test_message():
+            with pytest.raises(ImportError, match='Install "pkg" to use test_function'):
+                foo('test_function')
+
+    Source:
+        https://stackoverflow.com/questions/60227582/making-a-python-test-think-an-installed-package-is-not-available
+    """
+    import_orig = builtins.__import__
+
+    def mocked_import(name, *args, **kwargs):
+        if name == "pkg":
+            raise ImportError()
+        return import_orig(name, *args, **kwargs)
+
+    monkeypatch.setattr(builtins, "__import__", mocked_import)