Post

This reminds me of an exchange I had where I told a PR author (obviously an agent on the user's behalf) to review the AI Policy, and it actually closed the PR and apologized for violating the policy