How to handle spam / problematic use / bad bot behaviour?

[xurizaemon]

If a malicious user (or bot) interacted with @bolts in ways that contravene the CiviCRM Code of Conduct, are annoying or harmful to individuals / community, what could we do?

• Disable @bolts. I am happy to do this on request of CiviCRM core team, and @bolts could also be "locked out" by a password change by the core team. (I'm open to other requests, but I wanted this to be a concrete statement rather than something weasely like "by reasonable request", help me improve this wording?)
• Submitted a fix over in loafoe/hubot-matteruser#31 to make a configurable method to ignore users by name. So we have that feature now :)

[xurizaemon]

I've implemented a middleware for hubot, hubot-refrain

This lets hubot stop itself from saying certain things (based on patterns stored in robot.brain). So when we have had hashtagspam appear which contained text like "🆓 GET 66 FRIENDS 🆓", we are able to instruct bolts bolts refrain add GET .* FRIENDS and bolts will no longer voice those messages originating from hubot-twitter-mention which contain the matched string.