fix(filter): Allow generic Slackbot to pass

[BYK]

Fixes getsentry/sentry#14814

[BYK]

Eh, turns out this is not as straightforward as I thought. The generic bots? term eliminates the legit Slackbot 1.0 agent and the RegEx crate we use don't have lookaround support to only match when bot is not preceded with Slack.

The only two options I see are:

1. Remove the generic term (this has been there for 5+ years, scary)
2. Add a second layer that uses an allowlist and add Slackbot 1.0 there for by-pass

@jan-auer @untitaker thoughts?

[jan-auer]

Also adding @lobsterkatie since she's been adding filters over time. I have no context on implications of these filters.

For context, this was originally added due to getsentry/sentry#5284 in getsentry/sentry#5285

[jan-auer]

@BYK could you add more test cases for the bots we still want to block?

[BYK]

@BYK could you add more test cases for the bots we still want to block?

They are already there, that's why I didn't touch the tests much.

[lobsterkatie]

Happy to help but I have no special context on this. Over time in support, people would write in, asking for things to be added. I'd post in the engineering channel and get someone to give it a thumbs-up, and then make a PR adding to the file. It wasn't a super common thing, though.

[untitaker]

Wouldn't it be possible to enumerate linkexpanding and imgproxy instead?

[untitaker]

Answer: We have a generic bots filter that would match slackbot anyway.

[untitaker]

I think we won't get around having two regexes, but one last comment before we merge.