cc-switch

refactor(theme): remove circular reveal animation for theme switching

The View Transitions API used here crashes WebKitGTK with SIGSEGV on
Linux. Rather than gating document.startViewTransition per platform
(see PR #2502), drop the animation entirely — it's a low-value visual
flourish on a low-frequency action that doesn't justify a permanent
platform branch.

Removes the ::view-transition-* CSS block and the coordinate plumbing
in setTheme. The optional event parameter is kept on the API surface
to keep call sites compiling; they'll be cleaned up in a follow-up
commit.

Jason · 2026-05-02 20:33:54 +08:00

bc1f9341f4

fix(frontend): prevent selecting theme from causing segfault on Linux (#2502 )

Yuxuan Sun · 2026-05-02 17:27:11 +08:00

b61dad4b43

chore(usage): drop Hermes Agent tracking integration

Hermes aggregates all in-process API calls into a single sessions row
with the `model` field locked to the initial model, so the usage
dashboard cannot cleanly surface per-call billing context. Two rounds
of UI workarounds (raw mapping, then `<model> @ <host>` display) did
not resolve the user-facing confusion, so the whole tracking
integration is dropped for now.

Removes session_usage_hermes service (and its 17 tests), sync wiring
in commands/usage.rs and lib.rs, _hermes_session/hermes_session
entries in usage_stats SQL (provider_name_coalesce CASE and
effective_usage_log_filter IN clause), frontend Tab/banner/dropdown/
icon entries, and four i18n keys per locale.

Hermes app integration outside usage tracking (proxy routing,
session manager, config) is preserved. Pre-existing hermes rows in
proxy_request_logs are left as orphans — filtered out by the
updated SQL and never surfaced in the UI.

Jason · 2026-05-01 23:07:46 +08:00

518d945eb8

fix(claude): persist max effort via env (#2493 )

makoMakoGo · 2026-05-01 17:41:32 +08:00

064b339bab

feat(providers): add Baidu Qianfan Coding Plan for Claude Code (#2322 )

* feat(providers): add baidu qianfan coding plan presets

* refactor(providers): align qianfan presets with existing format

* chore(providers): narrow qianfan coding plan scope

jimmy · 2026-05-01 17:35:18 +08:00

db66348ff8

feat: support launch warp and execute session (#2466 )

* feat: support launch warp and execute session

Signed-off-by: tison <wander4096@gmail.com>

* other wires

Signed-off-by: tison <wander4096@gmail.com>

* for launch with provider

Signed-off-by: tison <wander4096@gmail.com>

* fixup indirection

Signed-off-by: tison <wander4096@gmail.com>

* clippy

Signed-off-by: tison <wander4096@gmail.com>

* address comments

Signed-off-by: tison <wander4096@gmail.com>

---------

Signed-off-by: tison <wander4096@gmail.com>

tison · 2026-04-30 10:33:50 +08:00

608ee35ecd

feat(usage): add Hermes Agent tracking + fix zero-cost bug + perf

Hermes:
- Parse ~/.hermes/state.db sessions (incl. profiles/*/state.db) into
  proxy_request_logs with data_source='hermes_session', WAL-aware
  incremental sync, Hermes-reported cost preferred over model_pricing
  fallback

Zero-cost bug (dashboard showed \$0 totals):
- GPT-5.5 family default pricing (~83% of affected rows used GPT-5.5)
- find_model_pricing_row: ASCII-lowercase normalization so
  "OpenAI/GPT-5.5@HIGH" matches seeded "gpt-5.5"
- Startup cost backfill in async task: scan rows where total_cost <= 0
  but tokens > 0, recompute via model_pricing in a single transaction

Performance:
- Add (app_type, created_at DESC) covering index for dashboard range
  queries
- Add expression index on COALESCE(data_source, 'proxy') so dedup EXISTS
  subqueries use index lookup instead of full scan; drop superseded
  idx_request_logs_dedup_lookup

Refactor:
- row_to_request_log_detail helper (3-way de-dup; fixes cost_multiplier
  \"1\" vs \"1.0\" drift between callers)
- Promote get_sync_state/update_sync_state to shared session_usage
  module (4 copies -> 1)
- run_step helper in lib.rs replaces 9 if-let-Err blocks
- maybe_backfill_log_costs returns bool to skip duplicate total_cost
  parsing in caller

Jason · 2026-04-29 22:41:46 +08:00

f061b777b7

chore(kimi): update Kimi For Coding website URL to /code/docs/

Sync the preset's websiteUrl from the legacy /coding/docs/ path to
the current /code/docs/ path across all four app presets (claude,
hermes, openclaw, opencode).

Jason · 2026-04-29 22:41:46 +08:00

bcf8434c1f

chore(codex): hide 1M context window toggle in provider edit form

Codex models no longer accept model_context_window=1000000, so the
toggle and its paired auto-compact-limit input are commented out in
the provider edit form. State hooks, helper imports, and i18n keys
are preserved so the UI can be restored in one batch if upstream
support returns. The TOML editor remains visible, allowing manual
edits if needed.

Jason · 2026-04-29 22:41:46 +08:00

5b6339d78c

feat(provider-form): soften validation with "save anyway" prompt (#2307 )

* feat(provider-form): soften business-rule validation with "save anyway" prompt

Refactor handleSubmit so empty-field / missing-item validations (provider
name, endpoint, API key, opencode model, template variables, provider key
required) no longer hard-reject with toast.error. Instead they are collected
into an issues list and presented via a ConfirmDialog; the user can cancel
or choose "Save anyway" to proceed.

Integrity constraints stay as hard rejections:
- providerKey regex / duplicate (would corrupt other providers)
- Copilot / Codex OAuth not authenticated (no token, cannot establish)
- omo Other Fields JSON not an object / parse failure

This aligns the frontend with the backend's existing "relaxed save / strict
switch" split (see gemini_config.rs: validate_gemini_settings vs
validate_gemini_settings_strict) and unblocks legitimate configs such as
AWS Bedrock, Vertex AI, and custom Gemini base URLs that the UI previously
refused to save.

Refs: #2196, #1204

* fix(provider-form): address review feedback on soft-validation

P1: move empty providerKey back to hard rejection for OpenCode / OpenClaw /
Hermes. Since providerKey is the primary identity for these apps and the
mutations layer throws "Provider key is required" when absent, letting users
click "save anyway" would surface a generic error toast instead of a
precise, actionable one. Treat empty providerKey as an integrity constraint
alongside regex / duplicate checks.

P2: give the soft-confirm submit path its own submitting state. The
confirm-dialog path bypassed react-hook-form's isSubmitting lifecycle, so
slow or failing saves left the outer submit button responsive and could
spawn unhandled rejections. Now the confirm handler awaits performSubmit
inside try/catch/finally, uses an isConfirmSubmitting flag to gate both
confirm and cancel clicks, and folds the flag into the outer disabled
state and onSubmittingChange callback.

Refs: #2307 review comments

* chore(clippy): use push for single char '…' in truncate_body

Clippy 1.95 added single_char_add_str which flagged the push_str("…")
in truncate_body. Rebased onto latest upstream/main and applied the
suggested fix so the Backend Checks clippy job passes.

Unrelated to this PR's core changes; bundled in so the PR is mergeable
without waiting for a separate upstream fix.

---------

Co-authored-by: Allen <allen@AllenMacBook-M4-Pro.local>

allen_xln · 2026-04-25 09:28:28 +08:00

85f0be9e1d

feat(compshare): add Coding Plan preset across claude/codex/hermes/openclaw

Introduce a dedicated "Compshare Coding Plan" variant pointing to
https://cp.compshare.cn (with /v1 for OpenAI-compatible apps). Reuses
the existing ucloud icon and promotion copy, while adding a new
providerForm.presets.ucloudCoding key in zh/en/ja.

Jason · 2026-04-24 23:57:04 +08:00

08e2b29b8a

feat(deepseek): switch presets to V4 (flash/pro) and add pricing

DeepSeek released V4 flash/pro; legacy IDs deepseek-chat / deepseek-reasoner
now alias to deepseek-v4-flash and will be deprecated.

- Update claude/hermes/opencode/openclaw presets to v4-pro / v4-flash,
  context 128K -> 1M; Claude Anthropic-compat endpoint routes OPUS/SONNET
  to v4-pro and HAIKU to v4-flash, plus an explicit modelsUrl override.
- Seed deepseek-v4-flash ($0.14/$0.28 per 1M) and deepseek-v4-pro
  ($1.68/$3.36 per 1M) into model_pricing; older v3.x / chat / reasoner
  rows kept for historical usage stats (INSERT OR IGNORE).
- Refresh user-manual (zh/en/ja) pricing table and note that legacy model
  IDs are billed at v4-flash rates.

Jason · 2026-04-24 23:24:37 +08:00

b1f9ce4653

fix(model-fetch): support /models for Anthropic-compat subpath providers

Providers like DeepSeek, Kimi, Zhipu GLM and MiniMax expose the
Anthropic-compatible API on a subpath (e.g. /anthropic) while the
OpenAI-style /models endpoint lives at the API root. The previous
heuristic blindly appended /v1/models to the Base URL, so every such
provider returned 404 and the UI mislabeled it as "provider does not
support fetching models".

Backend now generates a candidate list and tries them in order:
preset override -> baseURL /v1/models -> stripped-subpath /v1/models ->
stripped-subpath /models. Non-404/405 responses (auth, network) stop
immediately so we never retry against hostile status codes. Known
compat suffixes are kept in a length-descending constant so the
longest match wins; response bodies are truncated to 512 chars to
avoid HTML 404 pages bloating the error string.

Preset type gains an optional modelsUrl (DeepSeek points at
https://api.deepseek.com/models). Frontend threads the override
through fetchModelsForConfig when the current Base URL still matches
the preset default. A new fetchModelsEndpointNotFound i18n key
replaces the misleading "not supported" toast for exhausted-candidate
and 404/405 cases (zh/en/ja).

Jason · 2026-04-24 23:20:10 +08:00

67dbfc0a8c

fix(copilot): resolve Claude model IDs against live /models list

Copilot upstream returns model_not_supported when the client sends
dash-form Claude IDs (claude-sonnet-4-6, claude-sonnet-4-6[1m]) while
/models only accepts dot form (claude-sonnet-4.6, -1m suffix).

- Add copilot_model_map: syntax normalize (dash->dot, [1m]->-1m) plus
  live /models exact match and family-version fallback, reusing the
  existing 5 min auth cache. Returns None when the whole family is
  absent so upstream surfaces an explicit error instead of silently
  switching families.
- Wire into forwarder Copilot hook; runs before anthropic_to_openai
  conversion.
- Default Opus slot in the Copilot preset maps to Sonnet 4.6: Pro
  dropped all Opus on 2026-04-20 and Pro+ bills Opus 4.7 at 7.5x.
  Users who want real Opus can switch manually in the UI.

Refs: https://github.com/farion1231/cc-switch/issues/2016

Jason · 2026-04-24 22:58:58 +08:00

fcd83ee30d

feat(tray): show coding-plan usage for Kimi / Zhipu / MiniMax

dc04165f surfaced tray usage badges for Claude/Codex/Gemini official
OAuth only. Chinese coding-plan providers already expose 5h + weekly
windows through coding_plan::get_coding_plan_quota, but two gaps kept
the tray from rendering them.

- format_script_summary read only data.first(), truncating the tier-
  flattened UsageResult to a single window. Detect plan_name matching
  TIER_FIVE_HOUR / TIER_WEEKLY_LIMIT and emit the "🟢 h12% w80%" layout
  used by format_subscription_summary; worst utilization drives the
  emoji. Copilot / balance / custom scripts keep the legacy single-
  bucket output via fallback.

- usage_script previously required manual activation through
  UsageScriptModal. Auto-inject meta.usage_script on Claude provider
  creation when ANTHROPIC_BASE_URL matches a known coding plan, so the
  tray lights up without the user opening the modal. Does not overwrite
  existing usage_script on update.

Extract the URL route table out of UsageScriptModal into a shared
codingPlanProviders module so the modal, the creation hook, and the
Rust coding_plan::detect_provider mirror all agree on one list.
Add TIER_WEEKLY_LIMIT alongside TIER_FIVE_HOUR and a createUsageScript()
factory to collapse the duplicated default fields across four call
sites and drop the remaining stringly-typed tier names.

Jason · 2026-04-23 17:34:39 +08:00

4737158439

refactor(hermes): drop config health check scanner

The Hermes config.yaml schema has stabilized and users have migrated to
the current provider fields, so the value of scanning for model.provider
dangling references, custom_providers shape errors, v12 migration residue
etc. no longer justifies the maintenance surface — and the scan produces
false positives when users keep some providers under Hermes' v12+
providers: dict (Hermes' runtime merges both shapes, but CC Switch's
scanner only looked at the list form).

Removes the whole HermesHealthWarning type, scan_hermes_config_health
command, HermesHealthBanner React component, useHermesHealth hook,
warnings field on HermesWriteOutcome, and the three helper functions
(yaml_as_non_empty_str, collect_mapping_string_keys, hermes_warning)
that only served the scanner. Drops the matching i18n keys in
zh/en/ja and the fixInWebUI button label that only the banner used.

Jason · 2026-04-23 16:18:38 +08:00

cdcc423122

feat(tray): show cached provider usage in the system tray menu (#2184 )

* feat: add Rust-side write-through usage cache

Introduce an in-memory UsageCache on AppState that the existing usage
query commands populate on success. The cache is read-only to the rest
of the app today; the next commit consumes it from the tray menu.

- New services::usage_cache module with split maps: subscription keyed
  by AppType, script keyed by (AppType, provider_id).
- AppType gains Eq + Hash so it can be used as a HashMap key.
- commands::subscription::get_subscription_quota now takes State<AppState>
  and writes through on success (signature change is invisible to the
  frontend — Tauri injects State automatically).
- commands::provider::queryProviderUsage body extracted into an inner
  async fn; the public command wraps it with write-through, covering
  Copilot, coding-plan, balance, and generic script paths uniformly.

Cache is in-memory only; auto-query interval and the upcoming tray
refresh action rebuild it after restarts.

* feat(tray): surface cached usage in the system tray menu

Read UsageCache populated by the previous commit and render it in three
places, scoped to whatever TRAY_SECTIONS covers (Claude/Codex/Gemini):

1. Inline suffix on each provider submenu item
   "AnyProvider  · 🟢 5h 18% / 7d 23%"
2. Disabled summary row per visible app under "Show Main"
   "Claude · Anthropic Official · 🟢 5h 18% / 7d 23%"
3. "Refresh all usage" menu item that triggers get_subscription_quota +
   queryProviderUsage for every applicable provider, then rebuilds the
   tray menu via the existing refresh_tray_menu path.

Color encoding uses emoji (🟢 <70% / 🟠 70-89% / 🔴 ≥90%) since Tauri 2
tray labels are plain text. Missing cache entry leaves the label
unchanged — tray never issues network requests when opened. Three new
i18n-ready strings live in TrayTexts (en/zh/ja), following the existing
pattern for tray text.

Closes #2178.

* feat(usage): bridge tray UsageCache writes to frontend React Query

Why: tray hover triggers backend-only refresh that wrote to UsageCache but
never notified the frontend, leaving main UI stale while tray showed fresh
numbers. Emit a payload-carrying event after each cache write so React Query
can setQueryData directly, keeping both views in sync without duplicate fetches.

* fix(tray): skip hidden apps on hover refresh and drop stale disabled-script cache

Address P2 findings from automated review on #2184:

1. refresh_all_usage_in_tray now filters TRAY_SECTIONS by settings.visible_apps
   before scheduling subscription/script queries, matching create_tray_menu and
   preventing wasted external API calls (and rate-limit/auth-error log noise)
   for apps the user has hidden.

2. format_usage_suffix only trusts the script cache when provider.meta.usage_script
   is still enabled; when a script is disabled/removed the cached suffix is now
   invalidated so the tray label no longer shows stale data indefinitely.

* refactor: consolidate codex provider helpers and fix test semantics

- Add Provider::is_codex_oauth() and Provider::codex_fast_mode_enabled()
  to eliminate duplicated meta extraction in claude.rs and stream_check.rs
- Fix non-codex-oauth tests to pass codex_fast_mode=false (was true, harmless
  but semantically misleading)
- Remove redundant is_dir() guard after resolve_skill_source_dir already
  guarantees the returned path is a directory

* style: apply cargo fmt

* fix(tray): reflect failed refreshes in cache and support Gemini flash-lite

Follow-up to the tray usage-display feature addressing review feedback:

- Write snapshots for both Ok(success:false) and Err paths in
  queryProviderUsage / get_subscription_quota so stale success data
  no longer persists across failed refreshes; the original Err is
  still returned to the frontend onError handler.
- Include gemini_flash_lite tier in the tray summary with label "l".
  Matches the frontend SubscriptionQuotaFooter and keeps the worst
  emoji correct when lite is the highest utilization.
- Add TIER_GEMINI_PRO / _FLASH / _FLASH_LITE constants in
  services/subscription.rs and reuse them in classify_gemini_model
  and sort_order.
- Extract Provider::has_usage_script_enabled() to remove the
  duplicated meta.usage_script chain at two call sites.
- Use db.get_provider_by_id in refresh_all_usage_in_tray instead of
  materialising the full provider map, and parallelise subscription
  and script futures via futures::future::join.
- Narrow refresh_all_usage_in_tray to each section's effective
  current provider (script if enabled, else subscription when the
  provider is official). Hover refreshes now issue at most
  TRAY_SECTIONS.len() outbound requests.
- Add 10 unit tests in tray::tests covering Claude/Codex h/w dispatch,
  Gemini p/f/l dispatch (including lite-only and lite-worst cases),
  and success/failure guards.

---------

Co-authored-by: Jason <farion1231@gmail.com>

涂瑜 · 2026-04-23 16:17:15 +08:00

dc04165f18

Fix/一键配置失效 (#2249 )

* style(FailoverQueueManager): 显示供应商备注信息

* style(FailoverQueueItem): 添加供应商备注字段以支持备注信息显示

* style(FailoverQueueManager): 显示供应商备注信息

* style(FailoverQueueItem): 添加供应商备注字段以支持备注信息显示

* style(FailoverQueueManager): 更新供应商备注信息的显示样式

* style(FailoverQueueItem): 添加条件序列化以优化供应商备注字段

* fix: 优化模型状态管理，确保配置更新时正确引用最新设置

* fix(skill): improve error handling for skill source directory resolution

Co-authored-by: Copilot <copilot@github.com>

* fix(gemini): simplify project directory retrieval in scan_sessions function

* fix(useModelState): optimize latestConfigRef assignment in useModelState hook

* fix(useModelState): remove unnecessary blank line in useModelState hook

---------

Co-authored-by: Copilot <copilot@github.com>

Coconut-Fish · 2026-04-23 15:36:03 +08:00

010b163430

feat: Add Codex OAuth FAST mode toggle (#2210 )

* Add Codex OAuth FAST mode toggle

* fix(codex-oauth): default FAST mode to off to avoid surprise quota burn

service_tier="priority" consumes ChatGPT subscription quota at a higher
rate. Users must now opt in explicitly rather than inherit FAST mode
silently when this feature ships.

---------

Co-authored-by: Jason <farion1231@gmail.com>

Jesus Díaz Rivas · 2026-04-23 12:05:17 +08:00

10e0772d8c

fix(skills): prevent duplicate imports when import button is double-clicked (#2211 )

Closes #2139

Two related defects let the installed-skills count balloon when users
tap the import confirm button multiple times — either deliberately or
because the button is still clickable while a slow import is in flight:

- The confirm button only disabled itself while `selected.size === 0`,
  so it stayed clickable during a pending mutation. Each extra click
  triggered another `importFromApps` mutation.
- `useImportSkillsFromApps` appended the server response to the
  installed cache without deduping, so re-firing the mutation stacked
  the same skills into the list again.

Disable the confirm (and cancel) buttons while the mutation is pending
— matching the `isRestoring` / `isDeleting` pattern already used by
`RestoreSkillsDialog` — and merge success payloads by
`InstalledSkill.id` so repeated results overwrite rather than
accumulate.

The merge is extracted as a pure `mergeImportedSkills` reducer to make
the behaviour unit-testable and to short-circuit on an empty payload,
returning the existing reference so React Query does not notify
subscribers about a no-op cache update.

涂瑜 · 2026-04-23 10:25:01 +08:00

29e87487a1

Style/session manager list UI (#2201 )

* style(FailoverQueueManager): 显示供应商备注信息

* style(FailoverQueueItem): 添加供应商备注字段以支持备注信息显示

* style(FailoverQueueManager): 显示供应商备注信息

* style(FailoverQueueItem): 添加供应商备注字段以支持备注信息显示

* style(FailoverQueueManager): 更新供应商备注信息的显示样式

* style(FailoverQueueItem): 添加条件序列化以优化供应商备注字段

* style(App, SettingsPage, ScrollArea): 调整组件样式以改善布局和视觉效果

* style(App, SettingsPage, ScrollArea): 调整组件样式以改善布局和视觉效果

* style(SettingsPage, useSettings): 统一代码格式，调整样式和变量声明

* style(App): 调整底部内边距以改善布局

Coconut-Fish · 2026-04-23 08:56:57 +08:00

7bfe2609db

fix: always show name on hover (#2237 )

Signed-off-by: tison <wander4096@gmail.com>

tison · 2026-04-22 23:19:52 +08:00

1b3c53d235

feat(presets): upgrade Kimi K2.5 to K2.6 in direct Moonshot configs

Bump model id and display name from K2.5 to K2.6 in Hermes, OpenClaw,
OpenCode, and Claude (direct api.moonshot.cn) presets. Pricing,
context window, and base URL are unchanged.

Add kimi-k2.6 row to model_pricing seed; no migration needed since
seed_model_pricing uses INSERT OR IGNORE and runs on every startup
via ensure_model_pricing_seeded. Old kimi-k2.5 row is kept to
preserve historical usage stats.

Nvidia aggregator forwards (moonshotai/kimi-k2.5) intentionally keep
the K2.5 SKU until Nvidia's catalog confirms K2.6.

Jason · 2026-04-22 12:17:19 +08:00

1f25411524

feat(presets): add DDSHub Codex preset

DDSHub now exposes a Codex-compatible endpoint at the same host as
its Claude service. Base URL omits the /v1 suffix since the gateway
auto-routes OpenAI SDK paths.

Jason · 2026-04-22 10:42:24 +08:00

ae1fd2b662

feat(presets): add LemonData provider across all six apps

Register LemonData (third-party partner) preset for Claude, Codex,
Gemini, OpenCode, OpenClaw, and Hermes, plus icon assets and i18n
partner promotion copy (zh/en/ja). Claude preset uses
ANTHROPIC_API_KEY auth; OpenAI-compatible apps target gpt-5.4.

Jason · 2026-04-22 10:39:17 +08:00

f4a9021461

style(hermes): use dashboard icon and move MCP to toolbar end

Swap ExternalLink for LayoutDashboard on the Hermes Web UI button —
clicking it may launch `hermes dashboard` rather than just open a URL,
so a panel-style glyph reads truer than a generic external-link icon.
Also reorder the toolbar so MCP sits in the final slot, matching the
claude/codex/gemini/opencode layout.

Jason · 2026-04-21 23:05:52 +08:00

2274da0b8e

fix(usage): support Hermes and OpenClaw in usage query modal

Extend getProviderCredentials to read flat settingsConfig fields for
Hermes (snake_case base_url / api_key) and OpenClaw (camelCase baseUrl
/ apiKey), so the "official balance" template auto-selects for matching
providers like SiliconFlow.

Also refactor the BALANCE and TOKEN_PLAN test paths in handleTest to
reuse the precomputed providerCredentials instead of re-reading
env.ANTHROPIC_* directly, which previously caused empty key errors for
non-Claude apps even when the key was configured.

Jason · 2026-04-21 22:40:53 +08:00

4dd11ab427

feat(hermes): launch dashboard from toolbar when Web UI is offline

When the Hermes Web UI probe fails, the toolbar entry now opens an info
confirm dialog offering to run `hermes dashboard` in the user's preferred
terminal. Accepting spawns it via a temp bash/batch script; `hermes
dashboard` itself opens the browser once ready, so we do not poll.
The Memory panel and Health banner keep the existing toast behavior.

Also corrects the stale `hermes web` hint in the offline toast (the real
command is `hermes dashboard`) and reorders Linux terminal detection to
try `which` before stat'ing /usr/bin, /bin, /usr/local/bin.

Jason · 2026-04-21 21:11:15 +08:00

8e9452b7ec

style(providers): apply prettier to common-config hooks

Reformat three hooks brought in from upstream #2191 so that
format:check passes in CI.

Jason · 2026-04-21 12:05:48 +08:00

db92f2b3b0

fix(header): stop auto-compact from latching after maximize

useAutoCompact cached normalWidthRef = el.scrollWidth on every
non-compact resize, but per DOM spec scrollWidth === clientWidth
when content fits. Maximizing the window (content no longer
overflows) therefore wrote the container width into
normalWidthRef, making it impossible to re-enter compact when
the window was restored to its original size.

Move the assignment inside the overflow branch so the cache is
only written at the actual compact threshold, where scrollWidth
reflects the real content width.