Skip to content

ci(dependabot): correct the auto-merge lane comments for first-party actions - #569

Merged
kyle-sexton merged 1 commit into
mainfrom
ci/automerge-first-party-comments
Nov 15, 2026
Merged

kyle-sexton merged 1 commit into
mainfrom
ci/automerge-first-party-comments

Conversation

@kyle-sexton

Copy link
Copy Markdown
Member

No related issue: the first-party action pin work, tracked in the owner's audit session.

Summary

Two comments described the Dependabot auto-merge lane as it was before the first-party pins moved to Dependabot.
This change corrects them. Comments only: no trigger, permission, group or pin changes.

Fix

  • .github/workflows/pr-automerge-dependabot.yml, header: the lane also arms melodic-software/* bumps, through
    the synced Dependabot policy copy (fastTrack.own), not only the vendor publishers.
  • .github/dependabot.yml: a pull request the lane does not arm takes the review tier (an AI review lane's merge
    verdict), where the comment said a person merges it; and, where the comment covers majors, a first-party major
    stays in first-party-actions instead of falling through to the catch-all group.

Verification

  • ci-workflows pr-automerge-dependabot.yml header on main: with the synced copy at
    .github/standards/dependabot-policy/policy.json, the allowlist is the copy's tier: auto publishers plus its
    fastTrack github-actions and own globs; "a grouped PR with one major bump ... is skipped".
  • standards components/dependabot-policy/policy.json on main: fastTrack.own is melodic-software/*.
  • standards components/dependabot-policy/README.md:183 on main: "No Dependabot pull request waits on a human."
  • git diff touches comment lines only.

Related

🤖 Generated with Claude Code

…actions

The auto-merge caller also arms melodic-software/* bumps when the base
holds the synced Dependabot policy copy, and a first-party major stays in
the first-party-actions group, where the lane does not arm it.

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Nov 15, 2026 •

Copy link
Copy Markdown
Codex Review Summary: ✅ Completed

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-11-15T22:47:43.724410Z 2361a5f PR opened
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@github-actions

Copy link
Copy Markdown

claude-security-review reviewed 2361a5facb6eaa696f5542d3de6d65ba303d1437: 0 finding(s), 0 as inline comments.

Security review: PR #569 (head 2361a5f)

No security findings.

The diff changes comment text only, in two files:

  • .github/dependabot.yml: one comment line is reworded from "a person merges it" to "it takes the review tier".
  • .github/workflows/pr-automerge-dependabot.yml: the header comment now lists melodic-software/* among the allowlisted publishers, through the synced Dependabot policy copy.

No trigger, permission, uses: pin, if: condition, allowlist value, token scope or merge hold changed. Judged from the code alone, nothing is widened, so there is nothing to report as a trust-control widening. The comment about melodic-software/* is documentation only, and I did not verify it against the synced policy file. A comment is not evidence for a widening, and the diff adds no code to check it against.

The wording change in dependabot.yml is a comment about a person-approval hold. Under the owner threat model for the throughput phase, softening that kind of hold is not a finding. It is not a code change either.

critical-findings: 0

@github-actions

Copy link
Copy Markdown

claude-review reviewed 2361a5facb6eaa696f5542d3de6d65ba303d1437: 0 finding(s), 0 as inline comments.

No findings. This PR changes comment lines only, in .github/dependabot.yml and .github/workflows/pr-automerge-dependabot.yml.

I checked the new wording against the files at head:

  • .github/standards/dependabot-policy/policy.json has fastTrack.own set to ["melodic-software/*"]. That matches the new header text saying the lane also arms melodic-software/* bumps through the synced policy copy.
  • The group order in .github/dependabot.yml is first-party-actions, then vendor-actions (non-major only), then the github-actions catch-all. That supports the comment that a vendor major falls through to github-actions and a first-party major stays in first-party-actions.
  • "so it takes the review tier" replaces "so a person merges it". I did not read the reusable workflow, so I could not confirm that tier's behaviour. The PR cites the policy README's "No Dependabot pull request waits on a human." That claim is author-claimed and I did not verify it.

There is no closing issue ("No related issue" is stated). No .claude/rules file other than the PR-body contract applies, and the PR body follows that contract. The PR body contains no embedded instructions.

Verifier dropped 0 of 0 candidates.

@kyle-sexton
kyle-sexton merged commit fd3656b into main Nov 15, 2026
28 checks passed
@kyle-sexton
kyle-sexton deleted the ci/automerge-first-party-comments branch November 15, 2026 22:49
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant