Timo Kaufmann

The nagging message probably could use an overhaul with some more concrete actionable tips for the reviewer, since they might not be familiar with marvin. The usage is already explained...

Yes, I agree that this might make sense in the future. It could also potentially do things like pinging maintainers of relevant packages. That said, I probably will not work...

Because only the PR author can opt-in for now. You don't need to be part of the reviewer team to opt-in, but you do need to be the author.

Well in a way opt-in is restricted to PR authors precisely *because* Marvin is so new. I want to carve out some little safe corner of nixpkgs to experiment, while...

Discovered in https://github.com/NixOS/nixpkgs/pull/92697#pullrequestreview-447589745.

One possible fix might be to just re-fetch the issue and retry setting the status when label-removal fails. Alternatively, just give up at that point. Both options are not perfect,...

Hi! I think the main blockers are those that I listed as caveats [here](https://github.com/timokau/marvin-mk2/issues/34#issuecomment-732408066). It's a good idea to collect all of them here though. Off the top of my...

That would be great! I think the cleaning up step should come first though. We should at least double-check that nothing sensitive ends up in the log.

General idea: - Subscribe to [`check_run`](https://developer.github.com/webhooks/event-payloads/#check_run). - Reset state to `awaiting_changes` when a check [has conclusion](https://developer.github.com/v3/checks/runs/#get-a-check-run) negative or (debatable) has conclusion Neutral and is due to a ofBorg build failure....

Unfortunately the logs don't go far enough back, but I suspect that the bot actually requested a review of @ryantm 3 times and github only shows the first one. So...