feat(endpoints): upgrade to assistants v2

.github/workflows/pr.yml

@@ -0,0 +1,40 @@
+name: Tox
+
+on:
+  pull_request:
+    paths:
+      - "src/**"
+      - pyproject.toml
+
+jobs:
+  run:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: ["3.9", "3.10", "3.11", "3.12"]
+    steps:
+      - name: Checkout branch
+        uses: actions/checkout@v3
+        with:
+          ref: ${{ github.head_ref }}
+          fetch-depth: 0
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v4
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Setup Poetry
+        uses: abatilo/actions-poetry@v2.0.0
+        with:
+          poetry-version: 1.8.2
+
+      - name: Install dependencies
+        shell: bash
+        run: |
+          poetry config virtualenvs.in-project true && \
+          poetry install --with dev
+
+      - name: Tox run
+        shell: bash
+        run: poetry run tox run

.gitignore

@@ -1,5 +1,6 @@
 .mypy_cache/
 .pytest_cache/
+.tox/
 .venv/
 .vscode/
 __pycache__/

docs/endpoints/assistants/messages.md

@@ -1 +1,44 @@
 # Messages
+
+!!! warning
+
+    There is currently no support for attaching files. Subscribe to [#5: feat: add support for attached files for all assistants APIs](https://github.com/mharrisb1/openai-responses-python/issues/5) to be notified when this is added.
+
+!!! note
+
+    Only Assistants V2 is supported
+
+!!! tip
+
+    See [examples](https://github.com/mharrisb1/openai-responses-python/blob/main/examples/test_threads_api.py) for more
+
+## Decorator Arguments
+
+- `latency` - synthetic latency in seconds to introduce to the call(s). Defaults to `0.0`.
+- `failures` - number of failures to simulate. Defaults to `0`.
+- `state_store` - Optional [state store](../../user_guide/state.md) override for custom and shared states.
+- `validate_thread_exists` - Optional flag for asserting that thread exists. Defaults to `False`.
+
+## Create message
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## List messages
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Retrieve message
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Modify message
+
+!!! info
+
+    Docs are incomplete but feature is supported

docs/endpoints/assistants/run_steps.md

@@ -1 +1,30 @@
 # Run Steps
+
+!!! note
+
+    Only Assistants V2 is supported
+
+!!! tip
+
+    See [examples](https://github.com/mharrisb1/openai-responses-python/blob/main/examples/test_threads_api.py) for more
+
+## Decorator Arguments
+
+- `latency` - synthetic latency in seconds to introduce to the call(s). Defaults to `0.0`.
+- `failures` - number of failures to simulate. Defaults to `0`.
+- `state_store` - Optional [state store](../../user_guide/state.md) override for custom and shared states.
+- `steps` - Optional steps list. This is needed since there is no create steps route. Defaults to `[]`.
+- `validate_thread_exists` - Optional flag for asserting that thread exists. Defaults to `False`.
+- `validate_run_exists` - Optional flag for asserting that run exists. Defaults to `False`.
+
+## List run steps
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Retrieve run step
+
+!!! info
+
+    Docs are incomplete but feature is supported

docs/endpoints/assistants/runs.md

@@ -1 +1,58 @@
 # Runs
+
+!!! note
+
+    Only Assistants V2 is supported
+
+!!! tip
+
+    See [examples](https://github.com/mharrisb1/openai-responses-python/blob/main/examples/test_threads_api.py) for more
+
+## Decorator Arguments
+
+- `latency` - synthetic latency in seconds to introduce to the call(s). Defaults to `0.0`.
+- `failures` - number of failures to simulate. Defaults to `0`.
+- `state_store` - Optional [state store](../../user_guide/state.md) override for custom and shared states.
+- `sequence` - Optional state sequences for create and/or retrieve. Will use sequence `n` for call `n - failure`. Defaults to `{}`
+- `validate_thread_exists` - Optional flag for asserting that thread exists. Defaults to `False`.
+- `validate_assistant_exists` - Optional flag for asserting that assistant exists. Defaults to `False`.
+
+## Create run
+
+!!! warning
+
+    Messages created with `additional_messages` request params are currently ignored. Subscribe to [#10: fix: create run ignores `additional_messages`](https://github.com/mharrisb1/openai-responses-python/issues/10) to be notified when this is fixed
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Create thread and run
+
+!!! warning
+
+    Not implemented. Subscribe to [#11: feat: support create thread and run route](https://github.com/mharrisb1/openai-responses-python/issues/11) to be notified when support is added.
+
+## Retrieve run
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Modify run
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Submit tool outputs to run
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Cancel run
+
+!!! info
+
+    Docs are incomplete but feature is supported

docs/endpoints/assistants/threads.md

@@ -13,3 +13,27 @@
 - `latency` - synthetic latency in seconds to introduce to the call(s). Defaults to `0.0`.
 - `failures` - number of failures to simulate. Defaults to `0`.
 - `state_store` - Optional [state store](../../user_guide/state.md) override for custom and shared states.
+
+## Create thread
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Retrive thread
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Modify thread
+
+!!! info
+
+    Docs are incomplete but feature is supported
+
+## Delete thread
+
+!!! info
+
+    Docs are incomplete but feature is supported

docs/endpoints/index.md

@@ -20,7 +20,7 @@ Table is assembled according to APIs listed in the [OpenAI API Reference](https:
 | Moderations                            |     :material-close:     |          -          | Stateless    |
 | [Assistants](assistants/assistants.md) |   :material-check:[^3]   |          -          | Stateful     |
 | [Threads](assistants/threads.md)       |   :material-check-all:   |          -          | Stateful     |
-| [Messages](assistants/messages.md)     |   :material-check:[^3]   |  :material-close:   | Stateful     |
+| [Messages](assistants/messages.md)     |   :material-check:[^3]   |          -          | Stateful     |
 | [Runs](assistants/runs.md)             | :material-check:[^4][^5] |  :material-close:   | Stateful     |
 | [Run Steps](assistants/run_steps.md)   |   :material-check:[^6]   |          -          | Stateful     |
 | Vector Stores                          |     :material-close:     |          -          | Stateful     |

docs/index.md

@@ -5,7 +5,7 @@ Pytest plugin for automatically mocking OpenAI requests. See the [user guide](us
 !!! note
 
     This project does not try to generate fake responses from the models. Any part of a response that would be generated by a model will need to
-    be defined by the user or will be left as the fallback value (i.e. empty string for `str` type fields).
+    be defined by the user or will fallback to a default value.
 
 ## Installation
 
@@ -22,10 +22,3 @@ Available on [PyPi](https://pypi.org/project/openai-responses/)
     ```shell
     poetry add --group dev openai-responses
     ```
-
-## Why use this?
-
-For building production-ready AI features, you need to add tests to your code. A lot of work and attention is going into evals but the performance of the AI model is only part of the testing story.
-Your AI features live with all of the other code in your app and you need to be able to test those interactions.
-
-To avoid actually using the OpenAI endpoints in your tests - which cost money and are probabilistic - you can mock those calls so that everything works the way it will in production but with predetermined responses.

docs/user_guide/index.md

@@ -17,7 +17,7 @@ View available [endpoints](../endpoints/index.md).
 !!! note
 
     This project does not try to generate fake responses from the models. Any part of a response that would be generated by a model will need to
-    be defined by the user or will be left as the fallback value (i.e. empty string for `str` type fields).
+    be defined by the user or will fallback to a default value.
 
 ## Example
 

examples/test_threads_api.py

@@ -30,7 +30,6 @@ def test_create_thread_with_messages(threads_mock: ThreadsMock):
             {
                 "role": "user",
                 "content": "Hello, what is AI?",
-                "file_ids": ["file-abc123"],
             },
             {
                 "role": "user",

pyproject.toml

@@ -26,6 +26,7 @@ pytest = "^8.1.1"
 pytest-asyncio = "^0.23.6"
 types-decorator = "^5.1.8.20240310"
 mkdocs-material = "^9.5.18"
+tox = "^4.14.2"
 
 [build-system]
 requires = ["poetry-core"]

src/openai_responses/endpoints/_partial_schemas.py

@@ -8,6 +8,10 @@
 __all__ = ["PartialRun", "PartialRunStep"]
 
 
+class PartialIncompleteDetails(TypedDict):
+    reason: Optional[Literal["max_completion_tokens", "max_prompt_tokens"]]
+
+
 class PartialFunctionNoOutput(TypedDict):
     name: str
     arguments: str
@@ -34,18 +38,18 @@ class PartialLastError(TypedDict, total=False):
 
 
 class PartialRun(TypedDict, total=False):
-    status: RunStatus
-    required_action: PartialRequiredAction
-    last_error: PartialLastError
-    expires_at: int
-    started_at: int
     cancelled_at: int
-    failed_at: int
     completed_at: int
-    model: str
+    expires_at: int
+    failed_at: int
+    incomplete_details: PartialIncompleteDetails
     instructions: str
+    last_error: PartialLastError
+    model: str
+    required_action: PartialRequiredAction
+    started_at: int
+    status: RunStatus
     tools: List[AssistantToolParam]
-    file_ids: List[str]
 
 
 class PartialCodeInterpreterOutputLogs(TypedDict):