fix(ci): image tag was empty on scheduled runs #20

Merged
claude merged 1 commits from fix/image-name-on-schedule into main 2026-09-01 11:19:09 -04:00
Contributor

${{ github.event.repository.name }} resolves to empty on schedule events since the Gitea upgrade (now 1.27.0), so IMAGE became justin/ and the image build failed with invalid tag "192.168.0.2:1234/justin/:latest": invalid reference format — after the whole scrape + reindex had already succeeded.

  • IMAGE (and, where present, the metadata-action images: + OCI labels) now use ${{ github.repository }} — the run context, not the event payload. github.repository_owner from that same source was resolving fine all along.
  • Package-link / registry-GC steps take the bare repo name from ${GITHUB_REPOSITORY##*/} (POSIX, dash-safe).
  • New Verify image name resolves step fails in seconds rather than after a multi-hour scrape if this regresses.
`${{ github.event.repository.name }}` resolves to **empty** on `schedule` events since the Gitea upgrade (now 1.27.0), so `IMAGE` became `justin/` and the image build failed with `invalid tag "192.168.0.2:1234/justin/:latest": invalid reference format` — after the whole scrape + reindex had already succeeded. - `IMAGE` (and, where present, the metadata-action `images:` + OCI labels) now use `${{ github.repository }}` — the run context, not the event payload. `github.repository_owner` from that same source was resolving fine all along. - Package-link / registry-GC steps take the bare repo name from `${GITHUB_REPOSITORY##*/}` (POSIX, dash-safe). - New `Verify image name resolves` step fails in seconds rather than after a multi-hour scrape if this regresses.
claude added 1 commit 2026-09-01 11:12:14 -04:00
`${{ github.event.repository.name }}` resolves to EMPTY on `schedule`
events since the Gitea upgrade between 2026-07-01 and 2026-08-01, so
IMAGE became "justin/" and the monthly refresh died at the last step:

  ERROR: failed to build: invalid tag
  "192.168.0.2:1234/justin/:latest": invalid reference format

Both the 2026-08-01 and 2026-09-01 refreshes scraped, committed, pushed
and reindexed successfully, then failed to build the image — so the
corpus on main is current but the published container has been stale
since the 2026-07-01 run.

- IMAGE now comes from `${{ github.repository }}` (run context, not the
  event payload — `github.repository_owner` from the same source was
  resolving fine all along).
- The package-link and registry-GC steps take the bare repo name from
  `${GITHUB_REPOSITORY##*/}` (POSIX, safe under dash).
- New `Verify image name resolves` step fails in seconds instead of
  after a ~30 min scrape if this ever regresses.

Same fix applies to image-only.yml, which only escaped the bug because
push/workflow_dispatch events still carry the repository payload.

Co-Authored-By: Claude Opus 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01FnVuG79cYPcRLTp4pC8ujR
claude merged commit 953e0f10f9 into main 2026-09-01 11:19:09 -04:00
claude deleted branch fix/image-name-on-schedule 2026-09-01 11:19:09 -04:00
Sign in to join this conversation.