Repository navigation
chore(pstack): land card worktree commits and hold commits no ref reaches - #5182
Conversation
…ommit hold A card session in Plate may now commit on its worktree's own branch and lands everything since its merge base with the checkout branch, commits included, so committed card work no longer drops out of the patch. Cleanup holds a worktree whose HEAD history keeps a commit no branch, tag or remote reaches. Shipped fast, no review.
Review or Edit in CodeSandboxOpen the branch in Web Editor • VS Code • Insiders |
|
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
Why
A card session that committed in its own worktree lost that work at landing: the patch was
git diff --cached HEAD, which leaves out every commit. And the cleanup scan never checked whether a worktree's commits sat on any branch, so it could delete a worktree holding the only copy of a commit made on a detached HEAD.What changed
Synced from udecode/dotai@e23bf33, which carries ecaca2a and c13b58f:
git merge-base HEAD <checkout branch>, so the patch carries its commits, uncommitted edits and new files.--landedand--removehold a worktree whose HEAD or HEAD history keeps a commit that no branch, tag or remote reaches, and print thegit branch <name> <sha>that saves it.cleanupskill also saves a removal patch only for changes the checkout lacks.Verification
scan.mjs --landedpassed. The oldHEAD-based patch carried none of the commit.pnpm check:workflowpasses here.Shipped fast, no review.