[teamleaderleo]

New Tab's Search | Ask switch and its widget were already removed by #22674 and #22816. This PR finishes the config migration by retiring tabs.newTabWidgets.modeSwitch: an old file loads quietly, reports the retired key, and cannot expose it through Settings or agent writes.

The branch is current on feat-cmux-next base 3ad1a6bb71993c13e4ff9bd776d4fabd73ca41f3. Its complete base-to-head diff at 302a7cd279864cc2c2c87d70c138b6596c31cd2b is four settings files: retired-key source, generated Swift map, focused regression, and regenerated source stamp. Existing New Tab behavior and unrelated base changes are preserved.

Observed / Not run

• python3 scripts/cmux-next/gen-settings-sources.py --check: passed.
• python3 scripts/cmux-next/prepush-gate.py: passed for the own commits.
• python3 scripts/ci/defaults_guard.py check-stamp --base 3ad1a6bb71993c13e4ff9bd776d4fabd73ca41f3: passed.
• git diff --check: passed.
• Script tests (pin-cmux-tui, embed-cef, ensure-cef store) is base-red: bash scripts/cmux-next/tests/release-compile-derived-data.test.sh reproduces the bench OTHER_SWIFT_FLAGS failure on the untouched base worktree.
• Package dependency lists sorted is base-red: python3 scripts/cmux-next/check-package-dependency-lists.py reproduces the missing CmuxNextChiefMemory dependency on the untouched base worktree.
• Independent Codex xhigh re-review is running on the exact final head.
• The one exact-head PR wait is running at 302a7cd279864cc2c2c87d70c138b6596c31cd2b.
• Native builds/tests are not run locally; CI/fleet verification is authoritative.

No capture: this changes retired config-key handling only, with no layout or styling change.

Agent session: big-red

[teamleaderleo]

owner: 01a15c41-0d06-7430-b24d-a6d55b9be1d4

[teamleaderleo]

Human approval: Leo (teamleaderleo) approved this PR in the coordinator chat on 2026-10-20 ("i approve all 22612 22874 22883 22777"), at head 563072cf0956b764e18a595a19d0b5f29bb263ab. GitHub does not let the author account approve its own PR, so the approval is recorded here. A new push needs a new approval.

Agent session: cosmic-gray

[teamleaderleo]

Not merging at 563072c, the head Leo approved: it can't land as it is.

• Conflicts: git merge-tree against the base with cmux-next New Tab: a Minimal layout (greeting + composer) as the default, its extras as widgets #22674's current head (c5c9f6a) merged shows 563072c conflicting in 8 files: NewTabWidgetSettings.swift, new-tab-widgets.json, SettingsSchema+NewTabWidgets.generated.swift, SettingsSchema+AgentPolicy.swift, NewTabWidgetSettingsTests.swift, AgentPaneNewTab.swift, NewTabTemplateSetting.swift and docs/configuration.md. 563072c contains cmux-next New Tab: a Minimal layout (greeting + composer) as the default, its extras as widgets #22674 at 97aef40, and cmux-next New Tab: a Minimal layout (greeting + composer) as the default, its extras as widgets #22674 has changed since.
• Out of date: it also predates two newer cmux NEXT New Tab: open, ask or search from what you type, no switch #22816 commits (2679405 is cmux NEXT New Tab: open, ask or search from what you type, no switch #22816's head).
• Not reviewed: the independent review of this PR hasn't run yet.

Plan: once #22674 and #22816 merge, I'll merge the base in once, review the result, wait for a green Swift compile on that head, and post the new head here for a new approval.

[teamleaderleo]

Owned by nx-settings on big-red

[github-actions]

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

Review 302a7cd: 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=pr22883 -f ref=cmux/omnibox-auto-modeswitch -f platform=macos -f backend_mode=local

red: cmux-next checks (god files, concurrency, crash safety, l10n); no finished base run on 3ad1a6b to compare; do not merge until green; see lane-rules

[teamleaderleo]

Bead cx-104f3 created for this PR by the cmux coordinator.

[blacksmith-sh]

Found 11 test failures on Blacksmith runners:

Failures

Test	View Logs
/tmp/tmpi_mqbmzf/summary-missing.json (live):/summary-missing.json (live):	View Logs
/tmp/tmpiygc64p4/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
bench OTHER_SWIFT_FLAGS is 'OTHER_SWIFT_FLAGS=$(inherited) -DCMUX_BENCH_HOOKS -DCMUX_BL OCKING_CHECKS'/ bench OTHER_SWIFT_FLAGS is 'OTHER_SWIFT_FLAGS=$(inherited) -DCMUX_BENCH_HOOKS -DCMUX_BL OCKING_CHECKS'	View Logs
bench OTHER_SWIFT_FLAGS is 'OTHER_SWIFT_FLAGS=$(inherited) -DCMUX_BENCH_HOOKS -DCMUX_BL OCKING_CHECKS'/ bench OTHER_SWIFT_FLAGS is 'OTHER_SWIFT_FLAGS=$(inherited) -DCMUX_BENCH_HOOKS -DCMUX_BL OCKING_CHECKS'	View Logs
nightly cmux-tui resolve/nightly cmux-tui resolve	View Logs
test_the_nightly_is_covered (main.NightlyStepCoverageTests.test_the_nightly_is_cove red)/ test_the_nightly_is_covered (main.NightlyStepCoverageTests.test_the_nightly_is_cove red)	View Logs

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

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

[teamleaderleo]

Review (subagent, correctness first): SHIP at cb7ada1

• The generated SettingsSchema+RetiredKeys.generated.swift matches the generator (gen-settings-sources.py --check exits 0). No other copy of the retired-key list needs regenerating; cmux-tui reads the JSON directly.
• theRetiredModeSwitchKeyIsIgnored holds: the parse is unchanged, there is no diagnostic, retiredKeys reports the key, and there is no descriptor or agent-settable entry.
• Writes to the key are now refused as retired, which is what retiring means. No source code reads modeSwitch any more.
• Nit, not blocking: the AppServices comment lists only the agent and the model.

[teamleaderleo]

cmux-next crash-class rules / cmux-next crash-class rules is red on 9 of 9 fresh feat-cmux-next PR heads in the last 2:00:00, so it is most likely a base red, not this PR's. The base-reds lane has it; a base merge picks up the fix once it lands.

[teamleaderleo]

cmux-next guards / nightly-guards / workflow-guard-tests / release-notary is red on 10 of 14 fresh feat-cmux-next PR heads in the last 2:00:00, so it is most likely a base red, not this PR's. The base-reds lane has it; a base merge picks up the fix once it lands.

[teamleaderleo]

Review (subagent, correctness first): SHIP at 1fd5baa

• The new commit is the pr-autofix-patch artifact from run 39697000764 applied unchanged (its push job was skipped). It only rewrites settings-schema.sources.sha256.
• defaults_guard.py check-stamp exits 0 at this head. Every new digest matches the file's bytes.
• The full delta is still just the retired-key change plus the stamp.

[teamleaderleo]

Review (subagent, correctness first): SHIP at 59cf72e

• The second base merge's only conflict was the settings stamp. I took the base's stamp and updated only this PR's two sources to their source-stamp hashes. The lines the base leaves stale stay for the base autofix, so they keep reporting as base warnings instead of being marked fresh.
• check-stamp --base 390c286b exits 0 (only the base warning). gen-settings-sources.py --check exits 0.
• The delta from the base is identical to the earlier SHIP head 1fd5baa apart from the stamp. No base change was dropped.

[cursor]

Bugbot needs on-demand usage enabled

Bugbot uses usage-based billing for this team and requires on-demand usage to be enabled.

A team admin can enable on-demand usage in the Cursor dashboard.

[teamleaderleo]

Orphaned: no activity for 73 h, and no lane holds a lease on it. The orphan sweep could not merge feat-cmux-next in: conflicts in Packages/macOS/CmuxNext/Tests/CmuxNextSettingsTests/NewTabWidgetSettingsTests.swift, schemas/settings-export/settings-schema.sources.sha256. Listed for a lane to resolve.

[cursor]

Bugbot needs on-demand usage enabled

Bugbot uses usage-based billing for this team and requires on-demand usage to be enabled.

A team admin can enable on-demand usage in the Cursor dashboard.

[teamleaderleo]

cmux-next fast gates / post cmux-next/fast-gates is red on 18 of 23 fresh feat-cmux-next PR heads in the last 2:00:00, so it is most likely a base red, not this PR's. The base-reds lane has it; a base merge picks up the fix once it lands.

[teamleaderleo]

high-signal review link / link is red on 17 of 22 fresh feat-cmux-next PR heads in the last 2:00:00, so it is most likely a base red, not this PR's. The base-reds lane has it; a base merge picks up the fix once it lands.

[teamleaderleo]

Base-red attribution at source head 6c0d8870da97ec5b2662e6171ce3beefdcf6052c against base 7bd097da98a1a1d19277f43aea95ec032dce85e4:

The release-notary guard job fails on the release-note try-it requirement and existing nightly background-step/step-coverage/resolve guards. I reproduced those four Python test failures locally. Their workflows, scripts, release notes and test inputs are byte-identical to the base (git diff --exit-code over those paths passed), so these are inherited failures; Guard status inherits that job.

The settings source-stamp failure is this branch's and is being repaired through managed Mac autofix run 40262535390, with one condition wait active. I will apply its generated artifact and obtain a fresh exact-head review before landing.

[teamleaderleo]

Review (independent Codex /root/review_22883_final, xhigh): SHIP at db508d9

Independent final-head review found no blocking findings. The full delta is four files: the retired-key JSON, generated Swift map, focused NewTabWidgetSettingsTests regression, and the official one-line source-stamp artifact from managed Mac autofix run 40262535390. Source-stamp verification, generated-source verification, generated-output drift, JSON parsing and diff checks passed independently. The final repair changes only the generated stamp relative to the previously reviewed source head.

Validation limit: this was a read-only/static review; native Swift builds/tests were not run locally. I also inspected the final diff and clean worktree as the primary owner. Final-head PR checks remain in one condition wait.

[cursor]

Bugbot needs on-demand usage enabled

Bugbot uses usage-based billing for this team and requires on-demand usage to be enabled.

A team admin can enable on-demand usage in the Cursor dashboard.

[teamleaderleo]

Base-red attribution for live head 302a7cd279864cc2c2c87d70c138b6596c31cd2b against its base 3ad1a6bb71993c13e4ff9bd776d4fabd73ca41f3:

• Script tests (pin-cmux-tui, embed-cef, ensure-cef store) fails only because scripts/cmux-next/tests/release-compile-derived-data.test.sh expects bench OTHER_SWIFT_FLAGS without -DCMUX_BLOCKING_CHECKS; I reproduced the identical failure on the base worktree.
• Package dependency lists sorted fails because Packages/macOS/CmuxNext/Package.swift lacks CmuxNextChiefMemory in CmuxNextAppTests dependencies; python3 scripts/cmux-next/check-package-dependency-lists.py reproduces the same failure on the base worktree.

The PR's complete base-to-head diff remains four settings files only (retired-key source, generated Swift, focused regression, and source stamp), so neither red is caused by this PR. I made no unrelated fix.

[teamleaderleo]

Review (independent Codex /root/review_22883_final, xhigh): SHIP at 302a7cd

The final base-to-head diff is exactly the four intended settings files (19 insertions, 1 deletion). defaults_guard.py check-stamp --base 3ad1a6bb71993c13e4ff9bd776d4fabd73ca41f3, source-stamp parity, gen-settings-sources.py --check, generated-output drift, JSON parsing, and git diff --check all pass. No new findings.

The reported Script and Package dependency reds touch no changed path and reproduce unchanged on base, so they are inherited/base-owned. Validation was read-only/static; no native Swift builds or tests were run locally.

[teamleaderleo]

Final exact-head merge evidence for 302a7cd279864cc2c2c87d70c138b6596c31cd2b:

• Independent Codex xhigh review: SHIP at this exact head.
• cmux-next app compile (build fleet): SUCCESS (run 40268768402, job 121397015174).
• defaults guard, risk evidence, and merge freeze: SUCCESS.
• requiredMissing: empty; PR is MERGEABLE.
• cmux-next checks reds are inherited: Script tests and Package dependency lists reproduce unchanged on base 3ad1a6bb71993c13e4ff9bd776d4fabd73ca41f3.
• Defaults rerun failed only because GitHub installation API rate limit (HTTP 403), not a source failure.

[github-actions]

Shipped in:

• cmux NEXT nightly 1.0.0-nightly.4030749448501 (b781e3b4a0)