# 0hmx - https://github.com/tscircuit/tscircuit-a... Channel: #contributor Source: https://discord.com/channels/1233487248129921135/1519027866615025908 Started: 2026-06-23T17:14:26.255000+00:00 Last activity: 2026-06-24T04:24:12.737000+00:00 ## 0hmx — 2026-06-23T17:14:25.574000+00:00 https://github.com/tscircuit/tscircuit-autorouter/actions/runs/27930624946/job/82641645090 why dose this happen ? it auto fixes it self some times Attachment: image.png — https://community.tscircuit.com/media/1519027866589986977 ## 0hmx — 2026-06-24T04:10:09.880000+00:00 ``` • The run failed after the actual npm publish had already succeeded. Key points from the job: - Workflow: Publish to npm - Run: https://github.com/tscircuit/tscircuit-autorouter/actions/runs/27930624946 - Failing job step: Enable auto-merge - Package publish succeeded: @tscircuit/capacity-autorouter@0.0.602 - PR created: https://github.com/tscircuit/tscircuit-autorouter/pull/1428 - Failure line: GraphQL: Pull request Pull request is in clean status (enablePullRequestAutoMerge) What happened: the workflow ran gh pr merge 1428 --auto --squash --delete-branch. GitHub rejected enabling auto-merge because the PR was already considered clean/immediately mergeable at that moment. So the release side effects completed, but the bookkeeping PR never got auto-merge configured. Current state: - main is now at package version 0.0.606. - v0.0.602 tag exists. - PR #1428 is still open and now stale/conflicting. - Same bug later hit PR #1434 for v0.0.605, also still open/conflicting. - Later green publish runs don’t prove the bug is fixed; one green run skipped auto-merge because no new PR was created. Recommended cleanup: close stale PRs #1428 and #1434. Recommended workflow fix: handle the “clean status” case, either by directly merging when clean or by making the auto-merge step tolerate that specific GitHub error and use a fallback. ``` ## 0hmx — 2026-06-24T04:24:12.737000+00:00 https://github.com/tscircuit/tscircuit-autorouter/pull/1438