Skip to content

fix: prune skill backups when staging creation fails - #96

Merged
Upper9527 merged 1 commit into
mainfrom
codex/issue-triage
Oct 21, 2026
Merged

Upper9527 merged 1 commit into
mainfrom
codex/issue-triage

Conversation

@Upper9527

Copy link
Copy Markdown
Collaborator

When mirror pulls can create a backup but cannot create the staging directory, they still bypass retention after #88. Repeated failures can keep accumulating backups.

Move staging creation into the guarded operation and prune in finally. If staging creation fails, preserve the untouched local library instead of deleting it for an unnecessary restore. Keep the new backup available for recovery.

Add five regressions covering the #87 failure modes: dangling links, repeated partial backup failures, successful/failed rollback, and staging-directory failure. The original four fail before #88; the staging case fails on #88 and passes with this change.

Validation: 151 passed, 1 skipped (credential-dependent embedding integration); Ruff 0.12.2 lint/format and git diff --check pass. Local Python 3.11/macOS, temporary local storage, no live LLM calls.

Follow-up to #88 and #87.

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