The ArgoBEAST Lab (Experimental)

The ArgoBEAST Lab (Experimental)#

The ArgoBEAST Lab is a standardised, containerised testing environment. It provides a consistent execution space that includes a Selenium Grid and a dedicated ArgoBEAST Runner.

This is the recommended way to run tests in environments like WSL2, GitPod, or CI/CD pipelines where local browser management is often problematic.

Important

Important: The ArgoBEAST Lab is a fully isolated environment. It manages its own Python interpreter and dependencies internally. It is not compatible with host-managed virtual environments (e.g., .venv, uv, conda). If you use these tools locally, the Lab will ignore them in favour of its own internal “restocked” requirements.

Note

Note: The Lab is currently in an experimental phase. We are actively gathering feedback to refine its features and usability. If you encounter any issues or have suggestions, please reach out to us on GitHub or our community channels.

Commands Overview#

Command

Action

Description

argobeast build lab

Construct

Generates the Docker infrastructure and local directory.

argobeast open lab

Enter

Boots the Selenium Grid, updates config, and enters shell.

argobeast close lab

Sanitise

Shuts down the environment and frees up system resources.

1. argobeast build lab#

It is recommended you run this command in the early stages of your project setup, ideally right after argobeast init.

This command prepares the physical “equipment” for your testing environment. It creates a dedicated argobeast_lab/ directory in your project root.

  • Infrastructure: Generates a custom argobeast.dockerfile and argobeast.dockercompose.yml.

  • Isolation: Uses a Debian-slim base to ensure compatibility across WSL, Linux, and Mac.

  • Persistence: The Lab is built with an argouser to prevent file permission issues on your host machine.

Note

Run this once per project, or whenever you need to reset your Lab equipment.

2. argobeast open lab#

The primary gateway to your execution environment. When you run this command, ArgoBEAST performs a “System Pre-flight Check”:

  1. Integrity Check: Ensures Lab files exist.

  2. Config Sync: Automatically scans your config/driver.yml. If remote_url is missing or commented out, the Lab will wire it up for you.

  3. Engine Start: Spins up a Standalone Chrome container and the ArgoBEAST Runner.

  4. Infiltration: Executes an interactive session, placing you at the prompt: [argobeast lab]: /app #

Inside the Lab:

Once the doors swing open, you are in a pure Python environment. Simply run:

behave

to execute your tests. The Lab’s internal network allows seamless communication with the Selenium Grid, so your tests will run as if they were on a local machine.

You can run any argobeast command from within the lab, but remember that the Lab is designed to be a self-contained environment. If you need to make changes to your host machine’s configuration or files, exit the Lab first with:

exit

3. argobeast close lab#

When testing is complete, use this command to “turn off the lights.” It performs a docker compose down, ensuring no orphaned containers are eating your RAM.

Warning

For safety, you cannot close the Lab from inside the Lab. Type exit to return to your host machine first.

Technical Specifications#

The Selenium Grid (VNC)#

The Lab includes a built-in “Observation Window.” While the Lab is open, you can watch your tests execute in real-time:

  • URL: http://localhost:7900

  • Features: Live browser interaction and debugging via NoVNC.

Environment Variables#

The Lab injects the following into your session:

  • IS_IN_LAB=True: Used by the framework to prevent recursive loops.

  • ARGO_ENV=container: Can be used in your code to toggle specific behaviors.

  • SE_REMOTE_URL: Pre-configured to point to the internal Grid service.

Troubleshooting#

“The Lab is Locked” (Permission Denied)#

If you encounter a Docker socket error on Linux/WSL, your user needs permission to handle the equipment. Run:

sudo usermod -aG docker $USER
newgrp docker

“Missing Requirements”#

The Lab attempts to install your host’s requirements.txt upon build. If you add new dependencies to your project, you must run argobeast build lab again to “restock” the Lab’s libraries.