[refactor] chore: support ci mutli tests

.github/workflows/ci.yml

@@ -66,7 +66,7 @@ jobs:
         run: tools/ut.sh -j
 
   python_tests:
-    name: python tests (non-ollama) on ${{ matrix.os }} ${{ matrix.python-version}}
+    name: python tests (non-integration) on ${{ matrix.os }} ${{ matrix.python-version}}
     runs-on: ${{ matrix.os }}
     strategy:
       fail-fast: false
@@ -88,10 +88,10 @@ jobs:
         uses: astral-sh/setup-uv@v4
         with:
           version: "latest"
-      - name: Run Python Tests (excluding Ollama)
+      - name: Run Python Tests (excluding Integration)
         run: tools/ut.sh -p
         env:
-          PYTEST_SKIP_MARKERS: "ollama"
+          PYTEST_SKIP_MARKERS: "integration"
 
   ollama_integration_tests:
     name: ollama integration tests (linux python 3.10)
@@ -125,6 +125,31 @@ jobs:
           path: /tmp/ollama*.log
           retention-days: 7
 
+  integration_tests:
+    name: integration tests (linux python 3.10)
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Install java
+        uses: actions/setup-java@v4
+        with:
+          java-version: '11'
+          distribution: 'adopt'
+      - name: Install python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.10'
+      - name: Install uv
+        uses: astral-sh/setup-uv@v4
+        with:
+          version: "latest"
+      - name: Run Integration Tests
+        run: tools/ut.sh -p
+        env:
+          PYTEST_ONLY_MARKERS: "integration"
+          PYTEST_VERBOSE: "true"
+          CI: true
+
   e2e_tests:
     name: e2e tests on ${{ matrix.os }} ${{ matrix.python-version}}
     runs-on: ${{ matrix.os }}

python/flink_agents/integrations/chat_models/anthropic/tests/test_anthropic_chat_model.py

@@ -19,6 +19,9 @@
 
 import pytest
 
+# Mark all tests in this module as integration tests
+pytestmark = pytest.mark.integration
+
 from flink_agents.api.chat_message import ChatMessage, MessageRole
 from flink_agents.api.resource import Resource, ResourceType
 from flink_agents.integrations.chat_models.anthropic.anthropic_chat_model import (
@@ -27,7 +30,7 @@
 )
 from flink_agents.plan.tools.function_tool import from_callable
 
-test_model = os.environ.get("TEST_MODEL")
+test_model = os.environ.get("TEST_MODEL", "claude-3-haiku-20240307")
 api_key = os.environ.get("TEST_API_KEY")
 
 

python/flink_agents/integrations/chat_models/openai/tests/test_openai_chat_model.py

@@ -19,6 +19,9 @@
 
 import pytest
 
+# Mark all tests in this module as integration tests
+pytestmark = pytest.mark.integration
+
 from flink_agents.api.chat_message import ChatMessage, MessageRole
 from flink_agents.api.resource import Resource, ResourceType
 from flink_agents.integrations.chat_models.openai.openai_chat_model import (
@@ -27,7 +30,7 @@
 )
 from flink_agents.plan.tools.function_tool import from_callable
 
-test_model = os.environ.get("TEST_MODEL")
+test_model = os.environ.get("TEST_MODEL", "gpt-3.5-turbo")
 api_key = os.environ.get("TEST_API_KEY")
 api_base_url = os.environ.get("TEST_API_BASE_URL")
 

python/flink_agents/integrations/embedding_models/tests/test_openai_embedding_model.py

@@ -19,14 +19,17 @@
 
 import pytest
 
+# Mark all tests in this module as integration tests
+pytestmark = pytest.mark.integration
+
 from flink_agents.api.resource import Resource, ResourceType
 from flink_agents.integrations.embedding_models.openai_embedding_model import (
     OpenAIEmbeddingModelConnection,
     OpenAIEmbeddingModelSetup,
 )
 
 test_model = os.environ.get("TEST_EMBEDDING_MODEL", "text-embedding-3-small")
-api_key = os.environ.get("TEST_API_KEY")
+api_key = os.environ.get("TEST_API_KEY", "fake-key")
 
 
 @pytest.mark.skipif(api_key is None, reason="TEST_API_KEY is not set")

python/flink_agents/integrations/vector_stores/chroma/tests/test_chroma_vector_store.py

@@ -20,6 +20,9 @@
 
 import pytest
 
+# Mark all tests in this module as integration tests
+pytestmark = pytest.mark.integration
+
 try:
     import chromadb  # noqa: F401
 
@@ -35,9 +38,9 @@
     ChromaVectorStore,
 )
 
-api_key = os.environ.get("TEST_API_KEY")
-tenant = os.environ.get("TEST_TENANT")
-database = os.environ.get("TEST_DATABASE")
+api_key = os.environ.get("TEST_API_KEY", "fake-key")
+tenant = os.environ.get("TEST_TENANT", "fake-tenant")
+database = os.environ.get("TEST_DATABASE", "fake-database")
 
 
 class MockEmbeddingModel(Resource):  # noqa: D101

python/pyproject.toml

@@ -180,6 +180,7 @@ strict = true
 [tool.pytest.ini_options]
 markers = [
     "ollama: marks tests that require Ollama server (deselect with '-m \"not ollama\"')",
+    "integration: marks tests that require external services or integration (deselect with '-m \"not integration\"')",
 ]
 
 [tool.ruff.format]