Moderation is based on two things:
1. A spam/bad word filter. It's pretty aggressive and inspired by a variety of OSS tools, but definitely could be better. 2. An easy reporting tool available for any user-generated content. It could be more prominent for requests, so that's something I'll fix shortly.
I hacked together a small dashboard to surface whatever's caught by those two things. For the beta test, it's been working pretty well.
Feel free to ask any followups!