chore(deps): update taiki-e/install-action digest to 05a4690 - #394
Merged
Conversation
Dependency Review✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.OpenSSF Scorecard
Scanned Files
|
castrojo
approved these changes
Aug 6, 2026
castrojo
left a comment
castrojo
left a comment
Contributor
There was a problem hiding this comment.
Approved on behalf of @castrojo: taiki-e/install-action just tag currently resolves to commit 05a4690dfc1d29254f59ff1e818d7ef6bbfc7777, matching the pin. Earlier check failures were a transient GitHub 'Service Unavailable' during action download resolution; reruns are green.
castrojo
pushed a commit
that referenced
this pull request
Aug 6, 2026
Three defects found by auditing the gate against real merge behaviour observed on PRs #393, #394, #396, #397 and #383. 1. An absent check is not a passing check. `gh pr checks` reports only check-runs that already exist, so a queued workflow is missing from the rollup rather than PENDING — and a re-run removes its check-runs while they re-queue. Observed live on #394: three of seven checks were absent and the gate would have merged with them never having run. Because this workflow is triggered by one CI workflow completing while siblings may still be queued, that race is the normal case. Cross-check in-flight runs for the same commit via `gh run list --commit`. 2. An empty rollup or an unreadable run list now keeps waiting until the timeout instead of breaking immediately, so a transient API error is no longer terminal for a PR nothing else will re-evaluate. 3. Never infer the merge outcome from stderr or exit code. A plain `gh pr merge` enqueues while printing nothing and exiting 0, so the merge_method: queue path — the one used on a queued branch — reported every enqueue as "Merged". gh also writes its success line, including the PR title, to stderr, so a PR titled "...merge queue..." could be misread. Query the real PR state instead. All six gate scenarios re-verified, including the observed #394 and #397 cases; every ambiguous state now fails closed. Assisted-by: Claude Opus 5 via GitHub Copilot Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
6 tasks
castrojo
added a commit
that referenced
this pull request
Aug 6, 2026
* ci(actions): make mergeraptor auto-merge actually fire on main The mergeraptor review-bypass allowance was already live on main's branch protection, but nothing could use it: this repo had no workflow_run caller for reusable-renovate-automerge.yml, so qualifying Renovate PRs sat green and unmerged waiting on a human approval. Add the caller and harden the reusable workflow: - Add .github/workflows/renovate-automerge.yml, triggered by every CI workflow in the repo, passing base_branch: main, require_auto_merge: true and the MergeRaptor app credentials. - Mint a MergeRaptor installation token via app_id/private_key secrets. The bypass allowance only applies to that app identity, so github-actions[bot] could never have merged. Credential presence is mirrored into job-level env because the secrets context is unavailable in step-level if:. - Fix the author matcher. GraphQL reports an app author as the bare slug "mergeraptor" while REST reports "app/mergeraptor"; the previous gh pr list filter matched neither reliably. Normalise all spellings. - Gate merges on a complete check rollup. SKIPPED/NEUTRAL are non-blocking per GitHub's own merge semantics; PENDING is waited out (bounded by the new check_timeout_seconds input) rather than skipped, so a consumer that triggers from a single CI workflow cannot stall; anything else blocks. - Skip drafts and conflicting PRs, and let a failed merge surface as a job failure instead of a swallowed warning. Existing consumers (bluefin, bluefin-lts, dakota) pass only head_sha; both new inputs default to the previous behaviour, so this is additive for them. Assisted-by: Claude Opus 5 via GitHub Copilot Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs(skills): note that draft consumer PRs produce no CI The consumer validation protocol says to open a *draft* PR in projectbluefin/bluefin, but its PR Validation workflow never starts while the PR is a draft, so there is no run ID to cite as evidence. Pushing more commits does not help. Record the `gh pr ready` workaround. Assisted-by: Claude Opus 5 via GitHub Copilot Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): respect a merge queue when auto-merging `gh pr merge --squash` fails with "merge strategy for main is set by the merge queue" on a queued base branch — the queue owns the strategy. The hardcoded --squash would therefore have failed every merge on main. Add a merge_method input (squash by default, or "queue" to always omit the flag) and retry without the strategy flag automatically when that error is detected, so the workflow works on queued and unqueued branches alike. A genuine merge failure still fails the job. Assisted-by: Claude Opus 5 via GitHub Copilot Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): treat merge-queue advisories as success, not failure `gh pr merge --squash` on a merge-queue branch prints "! The merge strategy for main is set by the merge queue" and still enqueues the PR — with exit code 0. Exit status therefore says nothing about what happened; only the stderr text does. A retry can also hit "! Pull request <repo>#<n> is already queued to merge", which shares no substring with the first message and must be treated as success rather than a failed merge. Capture and echo stderr on every path, report enqueued vs merged accurately, and keep the no-flag retry for gh versions that hard-fail instead of warning. A genuinely unmergeable PR still fails the job. Verified against all eight real-world outcomes, using the verbatim strings observed while merging PRs #393, #396, #397 and #383. Assisted-by: Claude Opus 5 via GitHub Copilot Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): never treat an absent check or silent enqueue as success Three defects found by auditing the gate against real merge behaviour observed on PRs #393, #394, #396, #397 and #383. 1. An absent check is not a passing check. `gh pr checks` reports only check-runs that already exist, so a queued workflow is missing from the rollup rather than PENDING — and a re-run removes its check-runs while they re-queue. Observed live on #394: three of seven checks were absent and the gate would have merged with them never having run. Because this workflow is triggered by one CI workflow completing while siblings may still be queued, that race is the normal case. Cross-check in-flight runs for the same commit via `gh run list --commit`. 2. An empty rollup or an unreadable run list now keeps waiting until the timeout instead of breaking immediately, so a transient API error is no longer terminal for a PR nothing else will re-evaluate. 3. Never infer the merge outcome from stderr or exit code. A plain `gh pr merge` enqueues while printing nothing and exiting 0, so the merge_method: queue path — the one used on a queued branch — reported every enqueue as "Merged". gh also writes its success line, including the PR title, to stderr, so a PR titled "...merge queue..." could be misread. Query the real PR state instead. All six gate scenarios re-verified, including the observed #394 and #397 cases; every ambiguous state now fails closed. Assisted-by: Claude Opus 5 via GitHub Copilot Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Jorge Castro <jorge@castrojo.com> Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR contains the following updates:
6314fbb→05a4690Warning
Some dependencies could not be looked up. Check the Dependency Dashboard for more information.
Configuration
📅 Schedule: (UTC)
🚦 Automerge: Enabled.
♻ Rebasing: Never, or you tick the rebase/retry checkbox.
🔕 Ignore: Close this PR and you won't be reminded about this update again.
This PR has been generated by Mend Renovate.