The two leftovers I flagged in #966 and #968.
1. .config/buddy.ts named the wrong repository
repository: {
owner: 'stacksjs',
- name: 'development',
+ name: 'better-dx',Left over from before this repo was renamed. It has been working by accident rather than failing — GitHub redirects stacksjs/development to stacksjs/better-dx, so the bot's API calls landed in the right place anyway:
$ gh api repos/stacksjs/development --jq .full_name
stacksjs/better-dxSo this is hygiene, not an outage. Worth doing because the redirect stops applying the moment anyone creates a new repo under the old name, and because the config should say what it means.
To be clear about what this is not: the duplicate Dependency Dashboard issues in this repo predate the rename and were opened by chrisbbreuer and app/renovate, so they are unrelated to this setting.
2. publish:patch was byte-identical to release:patch
- "publish:patch": "bun x --bun @stacksjs/bumpx patch --recursive --commit --tag --push --yes",
+ "publish:patch": "bun run release:patch",After #968 added the flags, the two scripts were the same string, so they could drift apart silently. Delegating keeps the entry point working for anything that already calls publish:patch, rather than deleting it — happy to drop it instead if you would prefer, since bun-query-builder has no such script at all.
Verification
getConfig()returns{"owner":"stacksjs","name":"better-dx","provider":"github"}publish:patch's target script exists and resolves to the real commandpackage.jsonparses;bun --bun tsc --noEmitandpickierclean
I did not run release:patch or publish:patch — they commit, tag, push and publish.
🤖 Generated with Claude Code