Bug: fork of project fails action "issue-auto-unassign.yml" daily

[MarchandMD]

Impacted User Types

Developers

Environment

n/a

Current Behavior

Daily the issue-auto-unassign.yml runs and fails within my fork and I receive an email RE the workflow failure

Expected Behavior

ex: When I click "generate report", a downloadable report should be generated.

I don't expect to receive this failure daily (or ever) from my fork

How to Replicate

• Fork the repository
• Wait (usually occurs in the evening)
• observe workflow failure for issue-auto-unassign

How to access the QA site

Login Details:
Link to QA site

Login Emails:

• volunteer1@example.com view site as a volunteer
• supervisor1@example.com view site as a supervisor
• casa_admin1@example.com view site as an admin
• all_casa_admin1@example.com view site as an all casa admin
  • go to /all_casa_admins/sign_in

password for all users: 12345678

Questions? Join Slack!

We highly recommend that you join us in slack #casa channel to ask questions quickly. And discord for office hours (currently Tuesday 5-7pm Pacific), stakeholder news, and upcoming new issues.

[Sahancoder]

I’m also seeing this in my fork.

Every day the issue-auto-unassign.yml workflow runs in my fork and immediately fails, and I get the same notification email that all jobs have failed. I’m not actively using this workflow in my fork, so it’s just noisy for contributors who fork the repo.

From what I can tell, the problem is that the workflow is trying to perform write operations on issues, but the GITHUB_TOKEN in a fork doesn’t have permission to do that on the upstream repo, so the job fails.

One possible fix would be to skip this workflow when it’s running in a forked repository, for example:

jobs:
unassign:
if: ${{ github.event.repository.fork == false }}
runs-on: ubuntu-latest
steps:
# existing steps...

(or by checking `github.repository` against the upstream repo name).

If maintainers are happy with this approach, I’d be glad to open a PR to add the condition so forks don’t get these daily failure emails.

[MarchandMD]

@Sahancoder have you tried it locally? I imagine this fix would prevent the workflow from running locally; then, at least until this is merged to master/main, we'd have a workaround...