diff --git a/.github/instructions/copilot-instructions.md b/.github/instructions/copilot-instructions.md
new file mode 100644
index 000000000..0114e8f33
--- /dev/null
+++ b/.github/instructions/copilot-instructions.md
@@ -0,0 +1,15 @@
+---
+applyTo: '**'
+---
+
+## Indentation
+
+- Use 4 spaces for indentation.
+
+## Language
+
+- All comments must be written in English
+
+## Import Rules
+
+- Never import anything from other test files, a file will be treat as a module in pytest scope, so a new file will only able to use the shared fixtures in fixtures folder.
\ No newline at end of file
diff --git a/.github/instructions/harvester-test-instructions.md b/.github/instructions/harvester-test-instructions.md
new file mode 100644
index 000000000..bb98fd863
--- /dev/null
+++ b/.github/instructions/harvester-test-instructions.md
@@ -0,0 +1,119 @@
+## Harvester Test
+
+This instruction aims to explain how to write an E2E (end to end) integration and API test under [harvester_e2e_tests](../../harvester_e2e_tests/)
+
+- When adding a new Harvester API and the corresponding resource **does not already exist** under [Harvester API Manager](../../apiclient/harvester_api/managers/), create a new file with the **plural form** of the resource name.
+    - For example, if the resource is "noodle", search for `noodle` or `noodles` under [Harvester API Manager](../../apiclient/harvester_api/managers/). Refer to the [image API](../../apiclient/harvester_api/managers/images.py) or [storageclass API](../../apiclient/harvester_api/managers/storageclasses.py) examples
+    - Harvester API will be used in [integration tests](../../harvester_e2e_tests/integrations/) and [API tests](../../harvester_e2e_tests/apis/)
+    - Load the new resource [here](../../apiclient/harvester_api/api.py) and register it [here](../../apiclient/harvester_api/managers/__init__.py)
+    - Don't need to search [rancher API](../../apiclient/rancher_api/)
+
+- When writing **integration tests**, refer to the existing [tests directory](../../harvester_e2e_tests/integrations).
+    - Refer to the [images example](../../harvester_e2e_tests/integrations/test_1_images.py)
+
+- When writing **API tests**, refer to the existing [tests directory](../../harvester_e2e_tests/apis/).
+    - Refer to the [images example](../../harvester_e2e_tests/apis/test_images.py)
+
+- Each test case **must include at least one passing and one failing scenario**.
+
+- Follow these standard status codes to validate Harvester API responses:
+  - `201`: Resource successfully created
+  - `200`: Operation succeeded (general case)
+  - `422`: Operation failed 
+
+## Test Template
+
+Use this template when writing a new test.
+
+When calling API, don't use underline function or `create_data` function. If need, create another function or add more parameters to extend current function.
+
+### Shared Fixture Template
+
+```python
+@pytest.fixture(scope="module or class")
+def shared_fixture():
+    """
+    Shared fixture for test cases.
+    """
+    # 1. Prepare the environment and data
+    code, data = api_client.sss.xxx()
+
+    # 2. Assert the result
+    assert code == xxxx, f"Expected status code xxxx, got {code} with data: {data}"
+
+    output = {
+        "key1": "value1",
+        "key2": "value2",
+    }
+
+    yield output
+
+    # 3. Teardown if needed
+    api_client.sss.teardown_method()
+```
+
+### Test Case Template
+
+```python
+@pytest.mark.p0
+@pytest.mark.skip_version_if("???", reason="???")
+class TestXXXX:
+    """
+    Test description.
+    """
+
+    @pytest.mark.p0
+    def test_xxxx(self, ${some fixtures and shared_fixture}):
+        """
+        Test description.
+        """
+        # 1. Prepare the environment and data
+        shared_data = shared_fixture
+
+        # 2. Call the API
+        code, data = api_client.sss.xxx()
+
+        # 3. Assert the result
+        assert code == xxxx, f"Expected status code xxxx, got {code} with data: {data}"
+```
+
+## Harvester API Manager Template
+
+Use this template when implementing a new Harvester API manager.
+
+```python
+import base64
+from .base import BaseManager, DEFAULT_NAMESPACE
+
+
+class XxxsManager(BaseManager):
+    PATH_fmt = "/API/v1/namespaces/{ns}/xxx/{name}"
+    CREATE_fmt = "v1/harvester/xxx"
+
+    def create_data(self, name, namespace, data, annotations=None):
+        return {
+            "type": "xxx",
+            "metadata": {
+                "namespace": namespace,
+                "name": name,
+                "annotations": annotations,
+            },
+        }
+
+    def create(self, name, data, namespace=DEFAULT_NAMESPACE, annotations=None, **kwargs):
+        data = self.create_data(name, namespace, data, annotations=annotations)
+        return self._create(self.CREATE_fmt, json=data, **kwargs)
+
+    def get(self, name, namespace=DEFAULT_NAMESPACE, *, raw=False):
+        return self._get(self.PATH_fmt.format(ns=namespace, name=name), raw=raw)
+
+    def delete(self, name, namespace=DEFAULT_NAMESPACE, *, raw=False):
+        return self._delete(self.PATH_fmt.format(ns=namespace, name=name), raw=raw)
+```
+
+
+## Useful Fixtures
+
+### `ImageChecker`
+
+After a image is created or deleted, we need to check if it's done. So, we'll use the methods in [image_checker](../../harvester_e2e_tests/fixtures/images.py) to verfify it.
\ No newline at end of file