[teamleaderleo]

Leo, 2026-10-29: "Why the f*** did you remove the f***ing account switcher?"

What happened

#27151 ("Picker: choose Cmux as a provider, not an agent harness") renamed the composer model menu's Account row to Provider. Its order was "user should pick harness, and change provider to cmux"; nothing asked for the Account row to go. The account choices were still inside, but the row the composer had called Account (CLI login, Work Max, …) no longer said Account, so the switcher read as gone. Leo's later relayed ask (ask-067e9aa217) still wanted "accounts and variants inside" each harness.

Takes back #27151: only its renaming of the composer's Account row to Provider (and the gallery/test selectors that followed it), because Leo asked for the account switcher back; #27151's Cmux-as-provider choice stays, under a Provider heading inside Account.

What changed

• The composer model menu's account row is Account again (picker.account, already translated in all 21 locales), in the submenu and inline layouts.
• Cmux routers (cmux-router, local-coderouter) stay in that submenu, now under a Provider heading by default, so "change provider to cmux" still reads as a provider choice. Nothing is removed: the same rows, the same route callback, refusal handling and the model-offer Provider row (providers: "row") are unchanged.
• Gallery plays and the Playwright layout test look for the Account row again.

Test

webviews/test/model-menu-compact.test.tsx: new test "the composer keeps its Account switcher, with Cmux routers under a Provider heading", and the existing route tests open the Account row again.

• Red, test commit fb83d868da8 with base's menu: bun test test/model-menu-compact.test.tsx 33 pass, 8 fail (every Account-row test, including the new one: no row starts with "Account").
• Green, head 0e42761fbb5: 41 pass, 0 fail. src/agent-session/acpmux/QuickModelMenu.test.tsx 29 pass. vp fmt --check and vp lint clean on the changed files.

Risk

Low: a label and default grouping in one web menu. Any selector that looks for /^Provider/ on this row would miss; the repo's gallery plays and tests are updated.

Gallery: agent-pane.model-picker/started-chat-accounts, agent-pane.model-picker/quick-menu-many-profiles, new-tab.quick-model-menu/provider-submenu

Changelog

Fixed: the composer's model menu shows its Account switcher again; Cmux providers sit under a Provider heading inside it.

Agent session: agent session

🤖 Generated with Claude Code

---

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

---

Summary by cubic

Restores the composer model menu's Account switcher, which was unintentionally hidden when #27151 renamed the Account row to Provider.

• The Account row is back in the submenu and inline layouts, using the already-translated picker.account label.
• Cmux routers stay inside the Account submenu under a Provider heading, so choosing Cmux still reads as a provider choice.
• Gallery plays and Playwright layout tests target the Account row again.

^{Written for commit 4914802. Summary will update on new commits.}

[github-actions]

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

Review 4914802: 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=pr28020 -f ref=leo/next-composer-account-row -f platform=macos -f backend_mode=local

Gallery before/after for 4914802: 3 states rendered, 6 images differ (picked by the Gallery: line in the PR body). Images are in the run's artifact, not on this public PR; the independent review of this head reads them on its own, so the PR needs no hand-made screenshots of these states.

gh run download 41176797682 -R manaflow-ai/cmux -n gallery-advisory-28020-4914802939a457ca14bc3fe85fb13b1f77bd68a5 (then read gallery-review.json)

States

State	Theme	Status	Diff	Picked by
agent-pane.model-picker/started-chat-accounts	Apple System Colors Light	changed	0.131%	pr-body
agent-pane.model-picker/started-chat-accounts	Apple System Colors	changed	0.129%	pr-body
agent-pane.model-picker/quick-menu-many-profiles	Apple System Colors Light	changed	0.204%	pr-body
agent-pane.model-picker/quick-menu-many-profiles	Apple System Colors	changed	0.2%	pr-body
new-tab.quick-model-menu/provider-submenu	Apple System Colors	changed	8.17%	pr-body
new-tab.quick-model-menu/provider-submenu	Apple System Colors Light	changed	7.91%	pr-body

_{Pick other states with a Gallery: entry.id, entry.id/variant line in the PR body.}

Unclear: web / react-apps-check failed: no known signature; failed step: Check webview test shards.

CI failed on 4914802939 (run 41029315593 attempt 1): 2 code, 1 unknown.

Job	Verdict	Why
web / react-apps-base-check (1)	code	a web test failed
web / react-apps-base-check (3)	code	a web test failed
web / react-apps-check	unknown	no known signature; failed step: Check webview test shards

Failures:

• seen on other PRs pane-english.test.ts agent pane strings > no user-visible English is written in the pane's sources: failed (job)
• seen on other PRs composer-chip-press.test.ts chromium: a menu row takes a press anywhere on it, its check gutter and trailing edge included: failed (job)
• seen on other PRs composer-chip-press.test.ts webkit: a menu row takes a press anywhere on it, its check gutter and trailing edge included: failed (job)

Matched log lines

web / react-apps-base-check (1): (fail) agent pane strings > no user-visible English is written in the pane's sources [402.01ms]
web / react-apps-base-check (3): (fail) chromium: a menu row takes a press anywhere on it, its check gutter and trailing edge included [5936.98ms]

Not re-run automatically: web / react-apps-base-check (1), web / react-apps-base-check (3), web / react-apps-check are not machine failures.

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); NEW: red check: Package dependency lists sorted is not red on the base 8407795; do not merge until green; see lane-rules
red: web / Web status; base red: the same failure on the base 8407795; does not block
red: web / react-apps-base-check (1); no finished base run on 8407795 to compare; do not merge until green; see lane-rules
red: web / react-apps-base-check (3); no finished base run on 8407795 to compare; do not merge until green; see lane-rules
red: web / react-apps-check; base red: the same failure on the base 8407795; does not block

[blacksmith-sh]

Found 9 test failures on Blacksmith runners:

Failures

Test	View Logs
/tmp/tmp59vlbie7/summary-missing.json (live):/summary-missing.json (live):	View Logs
/tmp/tmpbo1kvbv1/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
test_every_listed_hook_is_one_glaeda_released_with_the_check (main.SigningHooksInSt epWithGlaedaTests.test_every_listed_hook_is_one_glaeda_released_with_the_check)/ test_every_listed_hook_is_one_glaeda_released_with_the_check (main.SigningHooksInSt epWithGlaedaTests.test_every_listed_hook_is_one_glaeda_released_with_the_check)	View Logs
test_the_repository_highlights_validate (main.DigestTests.test_the_repository_highl ights_validate)/ test_the_repository_highlights_validate (main.DigestTests.test_the_repository_highl ights_validate)	View Logs

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

[teamleaderleo]

Review (codex gpt-6-astra high, next-review): SHIP at 4914802

No blocking findings.

[cmux-bot]

cmux-bot approves 4914802: an independent review passed this exact head (next-review receipt by teamleaderleo). A push needs a new review; a later BLOCK at this head dismisses this approval.

[github-actions]

Shipped in:

• cmux NEXT nightly 1.0.0-nightly.4104058233901 (5a3b7e1cef)
• cmux NEXT nightly 1.0.0-nightly.4104292869901 (4e5fe0232d)
• cmux NEXT nightly 1.0.0-nightly.4104326854001 (e0f33500c8)
• cmux NEXT nightly 1.0.0-nightly.4104504184001 (afe5e8ddc4)