The problem was identified to be limited to a single user and has been resolved.

Users are experiencing occasional timeouts on hub.continue.dev. The incidence rate appears to be low, but we are investigating immediately