[teamleaderleo]

Runs the shared matrix's sidebar-order cases (Tests/Matrix/cases.json) headless on the app's own placement code. This covers cx-rq6i5, the follow-up to cx-gkyi1. The core runner (#20253) can't check these cases, because sidebar order is the home session's personal order: the app computes it and the daemon only stores it.

ModelMatrixPersonalOrderTests covers 15 cases: New Workspace Above, Below, At Top and At Bottom from the context menu, palette and CLI, plus Move to Bottom. For each case it:

• seeds three workspaces in personal order (for a create, also the new workspace, listed as its daemon lists it);
• runs the slot the action resolves to through WorkspaceVerbHandlers.place, the step every positioned create and Move to Bottom ends in;
• reads the sidebar's order straight away (the optimistic edit) and compares it with the case's order-is.

It runs once with personal-mixed-order-v1 and once without. It needs no daemon, window or clock, and every failure starts with MATRIX <id>:.

Not covered: Move Up/Down. movePersonalWorkspace only writes to the daemon and needs an active window, so a headless test sees no change. Those cases stay with the HID runner. Routing them through place would make them testable here, but it changes behavior, so it is not in this PR.

Verification

• Fleet run run-swift-testing-suites.sh on 7941bea: ModelMatrixPersonalOrderTests passed in both modes (0.23 s), and ModelMatrixTests passed.
• The one red in that run is DaemonCapabilityExportTests.exportIsFresh. plans/cmux-next/daemon-capabilities.json is stale on the base; this PR only adds one test file.

Changelog

none

🤖 Generated with Claude Code

---

^{Need help on this PR? Tag @codesmith-bot with what you need.}

---

Summary by cubic

Adds ModelMatrixPersonalOrderTests to run the shared matrix's sidebar-order cases against the app's own placement logic, which the core runner can't check because sidebar order is the home session's personal order. The test seeds three workspaces, runs each case through the app's placement code (creates via the pending-placements path, Move to Bottom via WorkspaceVerbHandlers.place), and checks both the sidebar's immediate order and the personal order written to the daemon, once with personal-mixed-order-v1 and once without. Move Up/Down cases remain with the HID runner because they only write to the daemon and require an active window.

^{Written for commit 4b54607. Summary will update on new commits.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

• Configuration used: Repository: manaflow-ai/cmux/.coderabbit.yaml
• Review profile: ASSERTIVE
• Plan: Advanced
• Run ID: 6232b6c3-6475-4e08-9318-79854941b909

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Not verified: macOS jobs did not run: compile and app tests were skipped on 4b54607697; ci-status is green without them.

CI passes on 4b54607697 (run 38492575027 attempt 1).
macOS jobs did not run: compile and app tests were skipped: Fast static checks skipped (the macos job needs both to succeed). Re-run CI once they pass, before merging.

Written by scripts/ci/classify_failures.py (ci-failure-attribution.yml); signatures are its SIGNATURES table. A machine verdict is the runner's fault, not this PR's; yours means the failing file is one this PR changes, also red on main that main's latest full suite fails the same way, seen on other PRs that it failed on another pull request's run lately.

[cursor]

Bugbot is paused — on-demand spend limit reached

Bugbot uses usage-based billing for this team and has hit its on-demand spend limit.

A team admin can raise the spend limit in the Cursor dashboard, or wait for the next billing cycle to continue.