Getting started
Install the controller and run your first test, step by step; the short form is Quick start. Complete the Prerequisites first.
Install the controller
Install from PyPI
If the controller machine is connected to the internet, install Orcraft from PyPI, in the activated virtual environment; the command is the same on every OS:
pip install orcraft
Install from a wheel file
If the machine has no internet access, download the wheel on a connected machine, copy it over and install it directly. A release carries one wheel per platform, so take the file for the machine's platform. Every wheel is built for Python 3.14's stable ABI, so the same file serves a later Python 3 as well. In the activated virtual environment:
| Platform | Command |
|---|---|
| Windows x64 | pip install orcraft-<version>-cp314-abi3-win_amd64.whl |
| Linux x86_64 | pip install orcraft-<version>-cp314-abi3-manylinux1_x86_64.manylinux_2_5_x86_64.whl |
| Linux ARM64 | pip install orcraft-<version>-cp314-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.whl |
| macOS, Apple silicon and Intel | pip install orcraft-<version>-cp314-abi3-macosx_10_15_universal2.whl |
Run the controller
orcraft-controller run
The first start after an install takes longer while the operating system checks the new files, up to about half a minute; the terminal counts the wait, and every later start is immediate (System reference/First start after an install).
Run the test
Clone the tutorial suite
Clone the Orcraft Tutorial repository, which contains a ready-to-run "Hello World" test:
git clone https://github.com/GMCoreAI/orcraft-test-suite-tutorial.git
Add the suite
- In the controller, click Add Item to add the tutorial suite to the session.
- Select the cloned tutorial suite's entry point Python file.
Select the test
Click on the Hello World test.
Start the run
- Click Run in the configuration panel on the right side.
- Confirm the run.
Check the result
Follow the test log as it executes.
A Pass result confirms that the Orcraft Test Framework is available for developing tests, and that local runs already work as expected.
If you're only interested in local runs, proceed to the Tutorial. Otherwise, continue to Adding endpoints to distribute test execution across machines.