We currently have documentation for using the REST API to imitate session activity to bump the workspace if a script is not working, but we have seen some reports of this not working.

Similarly, the coder schedule override-stop does not work if a policy is set on the template for autostop. Can we spend some cycles to have a revised opinion on how to recommend users keep the workspace running if they have, for example, a training job running?

Some ideas:

• Introduce a coder schedule bump CLI command
• Improved visibility in UI (a customer reported the autostop warning UI only shows up when the workspace hits the deadline, not always)
• verify the endpoint and/or update the docs