workspace backend API

[feloy]

The agents are executed by a workspaces backend.

(do we want the user to run workspaces in different backends, or each user can only choose one backend?)

Example of backends:

• podman
• docker
• apple containers
• microvm
• fake
• etc.

When a workspace is registered (kortex-cli init):

1. The instances manager calls the workspace backend, passing to the backend:

• name
• sources path
• config path

2. The backend returns:

• a unique ID

3. the instances manager registers the workspace with this ID

When a workspace is stopped (kortex-cli stop ID):

1. The instances manager stops the backend matching the ID
2. The instances manager unregisters the workspace

when the user wants to access the backend through ssh

1. the instances manager opens a connection between the user and the backend