At present what happens in CI / lifecycle.py is that all pull requests are checked and updated, which can cause chaos if the compute tag on the PR doesn't actually currently match what the QC Managers are looking for. It would be tidier to get the PR number from the triggering action (re-labelling) and only execute on that.
At present what happens in CI / lifecycle.py is that all pull requests are checked and updated, which can cause chaos if the compute tag on the PR doesn't actually currently match what the QC Managers are looking for. It would be tidier to get the PR number from the triggering action (re-labelling) and only execute on that.