Murilo Tavares

@DameonSmith IIRC you'd need to add `if: always() && needs.do-task-a.result == success` in this case. Once a job in the workflow is failed or skipped, you need to overwrite the...

Hi @niekvanderreest. In terms of the documentation, I haven't found that mentioned directly, but the docs do state: > A default status check of success() is applied unless you include...

That would be interesting. When a PR comes stale, there are labels that we want to remove automatically.

I see the same. But after some more debugging it seems the connection is fast, but it messes with github internals. What I did: I added steps before and after...

Code: ``` runs: using: "composite" steps: - name: print start date shell: bash run: | echo "Start date: $(date)" echo "START_TIME=$(date +%s)" >> $GITHUB_ENV - name: Tailscale uses: tailscale/github-action@v3 with:...

My theory is that the vpn breaks something in the github runner, which fails to report its conclusion. Probably that was done through an internal github route?

I am also wondering if one can use the exit node only for certain ip targets so that it doesn't mess with other routes

A quick update on this after dealing with it for a while: I'm pretty convinced it does NOT take 2 minutes as it looks like. At least in my scenario,...

> Seeing the same issue. Anyone know how to avoid breaking the GitHub Action view? @r-raymond I haven't but I learned to live with the limitations. if you encapsulate the...