We need to add flagging rules to the FAQ but off the top of my head, these are the valid reasons to flag:

• spam
• scam
• racism/bigotry
• threats of violence
• excessive and unnecessary meanness
• unwelcome nudity
• encouraging self-harm

A lot of these are subjective, which is why it's optional to see the consequence of flagging.

There is no "policy." Just suggestions, votes, payments, and algorithms.