I kind of rushed out #1733, reacting to some AI-generated ASCII art coming in.
However, I think we can define an AGENTS.md file that enforces our contributing guidelines and AI policy. I experimented with this repo, and found that AI agents (at least Copilot) do a pretty good job of following rules on what not to do: spenserblack/hostile-agents.md#1
So we could probably instruct them to never generate a PR, ensuring that a human has to at least create the PR and (hopefully) look over their contribution, and hopefully block fully automated PR spam submissions.
I kind of rushed out #1733, reacting to some AI-generated ASCII art coming in.
However, I think we can define an
AGENTS.mdfile that enforces our contributing guidelines and AI policy. I experimented with this repo, and found that AI agents (at least Copilot) do a pretty good job of following rules on what not to do: spenserblack/hostile-agents.md#1So we could probably instruct them to never generate a PR, ensuring that a human has to at least create the PR and (hopefully) look over their contribution, and hopefully block fully automated PR spam submissions.