It would be great to be able to bind cloud code to a system level error event. This would enable me send a request to an external service such as Slack to notify me of the error.

We have an 'Every Hour' script that runs and looks at the last hour of runtime.log entries and culls out tat are FATAL or ERROR level

(via "$or": [{"level": "FATAL"}, {"level": "ERROR"}]) and push the resulting list to our slack channels.  Could just as easily run every minute, I guess.