You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While reconciling the stale modeled-information-format-ci bypass-allowance drift (#80), mif-rs's live main branch protection was found to carry a different bypass pattern than every other checked repo:
GET /repos/modeled-information-format/mif-rs/branches/main/protection
.required_pull_request_reviews.bypass_pull_request_allowances.apps
-> [{"slug": "dependabot", "id": 29110, ...}] # the real, official GitHub Dependabot App
Every other repo checked (MIF, doc-site, structured-madr, ontologies) carried the stale modeled-information-format-ci entry (now reconciled to empty per #80) — none carried a dependabot bypass. mif-rs is the only repo seen with this shape.
Why it matters
This could be either:
Intentional — a deliberate per-repo choice to let the real dependabot App bypass review for mif-rs's own Dependabot PRs (consistent with the existing workspace convention captured in memory branch-protection-script-stale-app.md: "use dependabot slug + allow_auto_merge=true for repos with Dependabot auto-merge"), or
Note: a worktree already in this workspace (worktrees/dot-github-issue-667, branch fix/667-retire-stale-bypass-app, uncommitted) independently parameterizes branch-protection.sh with a BYPASS_APPS env var defaulting to "dependabot" — a third, still-uncommitted design that would make (1) the org-wide default rather than mif-rs-specific. That branch and #80's approach (empty allowance, no parameterization) are not reconciled with each other; whoever picks this issue up should resolve that inconsistency too rather than triage mif-rs in isolation.
Ask
A human (or a follow-up session with the mif-rs maintainer's context) should confirm whether mif-rs's dependabot-slug bypass is intentional and, if so, whether it should become the documented org pattern for repos with Dependabot auto-merge — reconciling with the BYPASS_APPS design already drafted in the fix/667-retire-stale-bypass-app worktree — or whether it should be reconciled to an empty allowance like the other four repos.
Repro
gh api /repos/modeled-information-format/mif-rs/branches/main/protection \
--jq '.required_pull_request_reviews.bypass_pull_request_allowances.apps[].slug'# -> dependabot
What
While reconciling the stale
modeled-information-format-cibypass-allowance drift (#80),mif-rs's livemainbranch protection was found to carry a different bypass pattern than every other checked repo:Every other repo checked (
MIF,doc-site,structured-madr,ontologies) carried the stalemodeled-information-format-cientry (now reconciled to empty per #80) — none carried adependabotbypass.mif-rsis the only repo seen with this shape.Why it matters
This could be either:
dependabotApp bypass review formif-rs's own Dependabot PRs (consistent with the existing workspace convention captured in memorybranch-protection-script-stale-app.md: "usedependabotslug +allow_auto_merge=truefor repos with Dependabot auto-merge"), orautomergeApp approves" pattern branch-protection.sh stamps the retired modeled-information-format-ci App into bypass_pull_request_allowances #80 establishes as the org norm.Note: a worktree already in this workspace (
worktrees/dot-github-issue-667, branchfix/667-retire-stale-bypass-app, uncommitted) independently parameterizesbranch-protection.shwith aBYPASS_APPSenv var defaulting to"dependabot"— a third, still-uncommitted design that would make (1) the org-wide default rather than mif-rs-specific. That branch and #80's approach (empty allowance, no parameterization) are not reconciled with each other; whoever picks this issue up should resolve that inconsistency too rather than triagemif-rsin isolation.Ask
A human (or a follow-up session with the mif-rs maintainer's context) should confirm whether
mif-rs'sdependabot-slug bypass is intentional and, if so, whether it should become the documented org pattern for repos with Dependabot auto-merge — reconciling with theBYPASS_APPSdesign already drafted in thefix/667-retire-stale-bypass-appworktree — or whether it should be reconciled to an empty allowance like the other four repos.Repro