Hostile-by-default critic that grades AI agent output against your acceptance policy.
Add this tool to a workspace, then choose which people and agents can use it.
No capability manifest has been published for this listing yet.