[teamleaderleo]

Reverts #23969 (squash c8f5d31). The tree is base with #23969's diff inverted exactly (7 files, +38/−348). Nothing else changes.

Why: #23969 merged on a same-model review. The different-model review that followed (Codex at xhigh) said BLOCK:

• Hidden while selected. OutboundTab.hides ignores shownByPane once a surface or name is marked. LinkOpener.move marks the tab before it awaits the move, so if the user selects that tab in the gap, the selected tab disappears from the strip. Selection, Cmd-W and tab indexes can then go stale.
• No departure check. finish(moved: true) releases the hold after 2 s regardless. If TabMoves acknowledges before the source pane's snapshot drops the tab, the tab reappears in the left strip, the same race cmux NEXT: a tab opened to the side starts its open on the right, never in the left pane #23969 set out to fix.
• Raw-id keys. Holds are keyed by tab id or surface alone, so an agent or provisional id reused while a hold is active is hidden too.
• Tests miss these. The tests mutate the snapshot directly and complete at once. They cover none of the cases above.

Before this PR: a tab opened to the side first shows on the right, with those edge cases. After: back to the behaviour before #23969, where the tab plays its open animation for about 100 ms in the left pane's strip and then moves right. A re-land with the fixes and a different-model review before merge follows.

Leo-ask: ask-f6988c508c
Leo-ask: ask-35d63f0002
Leo: "the animation for opening up a new tab starts off in the left pane, and it doesn't start off at the right"

No capture: an exact revert to the state before #23969; the before-and-after behaviour is described in #23969.

Agent session: cosmic-gray

🤖 Generated with Claude Code

---

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

---

Summary by cubic

Reverts #23969, restoring the previous behavior where a tab opened to the side plays its open animation in the left pane's strip for about 100 ms before moving right.

The revert is exact: #23969's diff inverted, so nothing else changes. It restores the state before that PR, which a different-model review blocked after merge for edge cases (OutboundTab hiding a tab a user selected mid-move, a 2 s hold released without confirming the source pane dropped the tab, and holds keyed by raw ids that can hide a reused id). The feature will be re-landed with those fixed and re-reviewed.

^{Written for commit 4caf7b2. Summary will update on new commits.}

[teamleaderleo]

Owned by main-pane-mode on cosmic-gray · session

[github-actions]

cmux status _{(one comment, edited in place; each line is updated by the job that owns it)}

Review 4caf7b2: HQ feed card · gallery before/after runs · checks

App of this head (app.zip, kept 1 day)

gh workflow run reload-build.yml -R manaflow-ai/cmux --ref feat-cmux-next -f tag=pr24292 -f ref=revert/23969-outbound-tab -f platform=macos -f backend_mode=local

Not verified: macOS jobs did not run: compile and app tests were skipped on 4caf7b27ed; ci-status is green without them.

CI passes on 4caf7b27ed (run 39743157947 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.

red: cmux-next checks (god files, concurrency, crash safety, l10n); no finished base run on b5a78a8 to compare; do not merge until green; see lane-rules

[teamleaderleo]

Review (subagent, correctness first): SHIP at 4caf7b2

Reviewer: Codex at xhigh, a different model from this lane's Claude. It is the same reviewer that blocked #23969.

• The tree is base with c8f5d31's diff exactly inverted. The 3-way revert applied cleanly on base b5a78a8.
• No OutboundTab references remain: git grep OutboundTab 4caf7b27 -- Packages/macOS/CmuxNext finds nothing.
• 4ef25f7 (cmux NEXT: revert #22822, links and pages from a chat never open as a tab in its pane #23886), which landed after cmux NEXT: a tab opened to the side starts its open on the right, never in the left pane #23969, does not change that.

The fleet compile-check --app passed at this head (job 104a257d).

[teamleaderleo]

Bead cx-wyia9 created for this PR by the cmux coordinator.

[blacksmith-sh]

Found 11 test failures on Blacksmith runners:

Failures

Test	View Logs
/tmp/tmp4_v6dkkj/summary-missing.json (live):/summary-missing.json (live):	View Logs
/tmp/tmp9hnyaima/summary-fail.json (live):/summary-fail.json (live):	View Logs
[new_workspaces] debug.new_workspace timed 0 of 50 new workspace rows/ [new_workspaces] debug.new_workspace timed 0 of 50 new workspace rows	View Logs
[new_workspaces] the window draws no frames/[new_workspaces] the window draws no frames	View Logs
[tabs] debug.tab_switch timed 0 of 40 switches/ [tabs] debug.tab_switch timed 0 of 40 switches	View Logs
[tabs] the window draws no frames/[tabs] the window draws no frames	View Logs
convention violations found (ERROR lines above)./ convention violations found (ERROR lines above).	View Logs
nightly cmux-tui resolve/nightly cmux-tui resolve	View Logs
test_every_submodule_free_checkout_on_an_owned_runner_resets_submodules (main.Reuse dWorkspaceSubmodules.test_every_submodule_free_checkout_on_an_owned_runner_resets_submo dules)/ test_every_submodule_free_checkout_on_an_owned_runner_resets_submodules (main.Reuse dWorkspaceSubmodules.test_every_submodule_free_checkout_on_an_owned_runner_resets_submo dules)	View Logs
test_every_submodule_free_checkout_on_an_owned_runner_resets_submodules (main.Reuse dWorkspaceSubmodules.test_every_submodule_free_checkout_on_an_owned_runner_resets_submo dules) (job='web-bundles')/ test_every_submodule_free_checkout_on_an_owned_runner_resets_submodules (main.Reuse dWorkspaceSubmodules.test_every_submodule_free_checkout_on_an_owned_runner_resets_submo dules) (job='web-bundles')	View Logs

...and 1 more test failures. View all on Blacksmith

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

[github-actions]

Shipped in:

• cmux NEXT nightly 1.0.0-nightly.3991367032401 (b9c3d5aabc)
• cmux NEXT nightly 1.0.0-nightly.3991575404801 (a62351daaa)

[github-actions]

Shipped in:

• cmux NEXT nightly 1.0.0-nightly.3991503232101 (cd84bc62d1)