The Knights appear undecided as to CLA signature status...

[doismellburning]

Apologies for this issue being a bit sparse on details but I don't have a lot of data I can provide...

In python/peps#851 the Knights correctly flagged it as CLA not signed

I subsequently submitted a signed CLA and checked back later with https://check-python-cla.herokuapp.com/ - this confirmed that I was registered as now having signed the CLA, and presumably kicked off a removal of the label by the Knights...only to have them reapply it immediately afterwards (see PR thread linked above)

Looks like the same thing happened to @brettcannon when he manually removed the label.

Funky internal state inconsistency / caching fun perhaps...?

[Mariatta]

The PR contains commits by two GitHub users: mindw and doismellburning.
The bot checks for CLA by everyone who made commits.in the PR.
Since mindw does not have bpo account, it applies the "CLA not signed" label.

To clear this, mindw can create the bpo account and sign the CLA, or create PR without commit by mindw.

[Mariatta]

I'm closing this issue since there is nothing to be fixed from the bot.

As mentioned, you have two options:

• Have the user @mindw to create a bpo account and sign the CLA
• Remove the commit made by @mindw (this commit python/peps@2807f75), and open a PR containing just your commit.

Thanks.

[doismellburning]

Right, thanks. Should I file a separate bug for "The Knights erroneously remove the label" or modify and reopen this one?

[Mariatta]

Removal of the label is not errornous, but rather a feature.
It is the simplest way for the bot to recheck the PR's CLA status, and it is how a core dev would have done it. I'm planning to move us to Cla-assistant soon. I don't think it's worth the effort to make changes to this now. Thanks.