It could first judge whether the PR is frivolous, then try to review it, then flag a human if necessary.
The problem is that Github, or whatever system hosts the process, should actively prevent projects from being DDOS-ed with PR reviews since using AI costs real money.
It could first judge whether the PR is frivolous, then try to review it, then flag a human if necessary.
The problem is that Github, or whatever system hosts the process, should actively prevent projects from being DDOS-ed with PR reviews since using AI costs real money.