* fix(markdown): render announcement markdown consistently
- support soft line breaks for announcement markdown without changing the default parser behavior.
- add explicit markdown element styles so lists, tables, code blocks, and quotes render correctly when typography styles are unavailable.
- apply the announcement markdown mode in both the popover and detail dialog for consistent display.
* refactor(markdown): simplify fallback markdown styles
- remove duplicate typography utility classes now covered by explicit markdown element fallbacks.
- keep the markdown renderer behavior unchanged while reducing class noise.
- modernize small helper expressions to satisfy targeted lint checks.
* fix(content): render custom HTML consistently
- add shared rich content rendering so custom HTML and Markdown use the same path across public pages and announcements.
- reuse common URL and HTML detection instead of duplicating content format checks per page.
- keep custom home content inside the standard public layout while preserving full-page iframe rendering for external URLs.
* refactor(playground): streamline chat request state
- extract conversation actions from the page component to keep message flow logic reusable.
- unify streaming and non-streaming generation state, including abort support for non-stream requests.
- simplify message rendering and payload construction while localizing Playground prompts.
* fix(playground): validate persisted chat state
- wrap saved Playground state with a storage version while still reading legacy values.
- validate config, parameter toggles, and messages before restoring them from localStorage.
- cap stored chat history to the latest messages to avoid oversized or stale state.
* refactor(playground): centralize message content access
- route chat rendering, copy actions, and error display through shared message helpers.
- reuse the current-version update helper for non-streaming assistant responses.
- keep message version details behind utility functions to reduce future model churn.
* refactor(playground): split storage schemas
- move Playground storage validation schemas into a dedicated module.
- keep storage read and write logic focused on migration, trimming, and persistence.
- preserve the existing storage envelope and validation behavior.
* refactor(playground): extract options loading hook
- move model and group queries into a dedicated hook so the page component stays focused on layout wiring.
- preserve existing fallback selection and error toast behavior while reusing the hook through the playground barrel export.
* refactor(playground): extract prompt suggestions
- move static prompt suggestion rendering into a focused component so the input stays centered on compose controls.
- preserve translated suggestion submission behavior while isolating icon metadata from the input form.
* refactor(playground): extract input tools
- move attachment and search controls into a dedicated component so the prompt input stays focused on compose state.
- keep existing development toast behavior and disabled handling while centralizing tool metadata.
* refactor(playground): extract input controls
- move model, group, send, and stop controls into a focused component so the input only manages compose state.
- preserve existing disabled states and generation button behavior while isolating control rendering.
* refactor(playground): extract message content display
- move sources, reasoning, loading, error, and response rendering into a dedicated message content component.
- keep the chat list focused on message iteration, edit state, and action wiring without changing display behavior.
* refactor(playground): extract message editor
- move inline message editing controls into a dedicated editor component so the chat list stays focused on rendering flow.
- preserve save, save-and-submit, cancel, and disabled-state behavior for edited messages.
* refactor(playground): extract stream error parsing
- move SSE error payload parsing into a reusable stream utility so the request hook stays focused on lifecycle handling.
- preserve existing error message, error code, and fallback behavior for raw or empty stream errors.
* refactor(playground): extract request error parsing
- move non-stream request error extraction into a shared utility so the chat handler stays focused on request flow.
- preserve the existing response message, error code, and fallback priority for failed chat completions.
* refactor(playground): extract streaming chunk updates
- move reasoning and content chunk application into a message utility so the chat handler only wires stream events.
- preserve error-state skipping, reasoning accumulation, and content streaming behavior for assistant messages.
* refactor(playground): extract message reasoning parser
- move think tag parsing into a dedicated playground message utility.
- export the parser through the shared playground lib barrel for consistent imports.
* refactor(playground): extract message streaming utilities
- move stream chunk application and message finalization into a dedicated utility.
- keep stored message sanitization with the streaming lifecycle helpers.
* refactor(playground): extract message update utilities
- move assistant message update helpers into a focused playground utility.
- keep error-state message updates separate from core message construction helpers.
* refactor(playground): extract completion choice handling
- move non-streaming choice application into the message streaming utilities.
- keep the chat handler focused on request orchestration and message updates.
* refactor(playground): centralize assistant completion state
- add a helper for finalizing assistant messages with complete status.
- reuse the helper in stream completion and stop-generation paths.
* refactor(playground): extract stream message parsing
- move SSE delta parsing into a shared stream utility.
- keep the stream request hook focused on lifecycle handling and update dispatch.
* refactor(playground): extract stream ready state checks
- move SSE ready-state status handling into stream utilities.
- keep weak source status typing outside the stream request hook.
* refactor(playground): extract conversation message helpers
- move send, regenerate, and edit message list construction into focused utilities.
- keep the conversation hook focused on edit state and update dispatch.
* refactor(playground): extract state initialization helpers
- move playground initial state loading into focused utility helpers.
- centralize message state updater resolution outside the React state hook.
* refactor(playground): extract option fallback helpers
- move model and group fallback selection into focused playground utilities.
- keep the options hook focused on query results, toasts, and config updates.
* refactor(playground): extract message action helpers
- move message action state derivation into focused utilities.
- keep the action component focused on guarded handlers and rendering.
* refactor(playground): extract input control state
- move submit, stop, and selector state derivation into a pure helper.
- keep input controls focused on rendering model selectors and action buttons.
* refactor(playground): extract message content state
- move source, reasoning, loader, and body visibility checks into a pure helper.
- use a discriminated state shape so rendered reasoning content stays type-safe.
* refactor(playground): extract message editor state
- move save eligibility and submit visibility checks into a pure helper.
- keep the editor component focused on textarea and button rendering.
* refactor(playground): extract message error state
- move error kind, fallback content, and admin visibility checks into a pure helper.
- centralize the model pricing settings path used by the error action.
* refactor(playground): extract chat render state
- move editing content lookup and per-message render flags into conversation helpers.
- keep the chat component focused on mapping messages to editor and content views.
* refactor(playground): extract suggestion display state
- move suggestion class selection into a pure helper.
- keep the suggestions component focused on translation and rendering.
* refactor(playground): extract assistant message state checks
- move final and pending assistant status checks into streaming utilities.
- keep the chat handler focused on request lifecycle updates.
* refactor(playground): extract input tool state
- move attachment action metadata and development notices into input tool utilities.
- keep the input tools component focused on menu and button rendering.
* refactor(playground): extract stream protocol checks
- move SSE done-message and closed-ready-state checks into stream utilities.
- keep the stream request hook focused on event handling flow.
* refactor(playground): extract message removal helper
- move delete-message filtering into conversation message utilities.
- keep the conversation hook focused on action orchestration.
* refactor(playground): extract option error messages
- move option load error message selection into playground option utilities
- keep the options hook focused on query effects and fallback updates
* refactor(playground): extract input submit text helper
- move prompt submit text validation into input control utilities
- let the input component submit only when a concrete text value is available
* refactor(playground): centralize error message checks
- add a shared helper for identifying error messages
- remove direct status string checks from message content rendering
* refactor(playground): extract message content display checks
- move loader and content visibility decisions into local helper functions
- keep message content state assembly focused on composing render state
* refactor(playground): replace raw message role checks
- use shared message role constants in conversation edit handling
- avoid raw assistant role literals when validating API messages
* refactor(playground): extract non-stream response handling
- move chat completion response choice handling into message streaming utilities
- keep the chat handler focused on request lifecycle and error routing
* refactor(playground): centralize stream cleanup
- reuse one stream cleanup path for completion, errors, startup failures, and manual stops
- preserve the current-source guard when closing SSE streams
* refactor(playground): extract pending assistant check
- centralize pending assistant message detection in streaming utilities
- reuse the helper when sanitizing stored playground messages
* perf(playground): improve mobile input controls
- split mobile input controls into selector and action rows
- keep the desktop input footer compact while reducing mobile control crowding
* perf(playground): add starter empty state
- show starter prompts in the empty playground chat area
- wire empty-state prompt selection into the existing send flow
- add localized copy for the new empty state
* perf(playground): improve mobile message actions
- collapse mobile message actions into a touch-friendly dropdown menu
- keep the desktop hover action strip unchanged for pointer workflows
- share one action list between desktop buttons and the mobile menu
* perf(playground): add error recovery actions
- show retry, edit, and delete actions inside error message alerts
- route edit recovery to the previous user prompt when available
- keep recovery controls touch-friendly on mobile layouts
* perf(playground): refine message editing experience
- present message edits in a focused bordered editor panel
- add unsaved-change state, reset, and cancel confirmation flows
- improve mobile touch targets and keyboard shortcuts for editing
* perf(playground): improve markdown code blocks
- render fenced markdown code with syntax highlighting, line numbers, and fallback plain text
- add copy, download, and collapse controls for playground AI responses
- tighten code block layout and theme token styles for responsive markdown rendering
* fix(playground): constrain markdown code block height
- collapse long playground code blocks after a short preview instead of waiting for very large snippets
- cap expanded code blocks so long responses scroll inside the code block
- keep generic code block usage unconstrained unless a caller opts in
* feat(playground): add chat history clearing
- add a toolbar action that is enabled only when saved playground messages exist.
- confirm destructive clears before removing browser-stored conversation state.
- add localized strings for the action, dialog, and completion toast.
* perf(playground): improve chat markdown rendering
- refine assistant and user message surfaces so chat content matches the app UI.
- normalize markdown typography, tables, images, lists, blockquotes, and details rendering.
- add indentation cues for collapsible reasoning and source sections.
* style: format code block component
* style: format playground frontend files
* feat(playground): render markdown with stream parser
- replace Streamdown with stream-markdown-parser for project-owned markdown rendering and styling.
- split response rendering into focused block, inline, table, alert, details, and footnote modules.
- pass message final state into response parsing so streaming content can be parsed incrementally.
* fix(playground): localize reasoning and chat feedback
- translate reasoning status, message actions, playground errors, and response renderer fallbacks across supported locales.
- keep reasoning duration numeric and tighten the collapsible layout to prevent trigger jitter.
- register dynamic keys so i18n sync keeps runtime labels covered.
* refactor(playground): group files by functional area
- move chat, input, and message components into focused subdirectories to make the UI structure easier to scan.
- split playground helpers into input, message, streaming, storage, options, state, and suggestions modules.
- update barrel exports and imports so existing feature entry points continue to work.
* fix(playground): prevent history replay from freezing page
- defer saved conversation loading so route entry no longer blocks on localStorage parsing and markdown rendering.
- limit initial history rendering and skip expensive markdown parsing for oversized responses.
- normalize corrupted streaming snapshots and cumulative chunks to keep saved playground history bounded.
- add message timing metadata and layout alignment groundwork without introducing live timers.
* feat(playground): allow regenerating from user messages
- show regenerate actions on user messages with saved content.
- truncate following conversation state before starting a fresh assistant response.
* feat(playground): add raw response source view
- add a per-message source toggle for assistant responses.
- render raw response content with the existing code block viewer.
- localize the new source and preview action labels.
* feat(playground): render code with unified editor
- replace Shiki HTML rendering with a read-only CodeMirror view for code blocks and raw responses.
- reuse the same CodeMirror frame for message editing so source and edit modes stay visually aligned.
- add lightweight CodeMirror dependencies while keeping language support scoped to Markdown.
* perf(playground): streamline chat input controls
- combine model and group selection into one compact picker for faster context switching.
- switch playground action buttons to icon-first controls with tooltips to reduce toolbar width.
- refresh input footer styling and submit states so active and destructive actions are clearer.
- bump dompurify lockfile entry to keep the frontend dependency current.
* fix(playground): filter models by selected group
- query user models by the selected playground group instead of reusing the cross-group model union.
- clear unavailable model selections and block sending when the active group has no models.
- align model selector and error action controls with the existing playground interaction style.
* perf(playground): remove input suggestion chips
- remove the prompt suggestion row below the playground input to reduce visual noise.
- delete the now-unused suggestion component and display helper.
* perf(playground): stabilize reasoning trigger layout
- use fixed icon slots around the reasoning label so the left content stays still when toggling.
- limit the open state animation to the chevron rotation for a smoother collapse interaction.
* perf(playground): smooth reasoning expansion
- use the collapsible panel height animation for vertical reasoning reveals.
- sync inner content opacity and position with the panel state.
* perf(data-table): autosize action columns
- exclude actions columns from shared table width calculations so action cells size to their content.
- remove fixed size and w-* width overrides from feature action columns to preserve content-based layout.
* perf(data-table): streamline row action controls
- expose common edit and status actions directly while moving secondary actions into overflow menus.
- add shared row action menu helpers so static and table rows use consistent action controls.
- let action columns size to their content instead of relying on fixed widths.
* fix(web): localize destructive dialog copy
- route delete, reset, and batch update confirmation text through i18n.
- add locale entries for affected channel, model, system settings, and user dialogs.
* perf(web): unify destructive dialog actions
- align delete and cleanup confirmation buttons with the shared destructive variant.
- replace custom destructive color overrides with semantic button variants.
- clean up lint errors in touched dialog files before committing.
* fix(web): add user action success translations
- add localized success messages for user delete, status, and role changes.
- keep user management toast copy available across all frontend locales.
* fix(data-table): prevent mobile badge clipping
- expose badge cell slots so mobile card styles can target nested badge wrappers.
- reset badge margins in card rows to keep provider icons fully visible on small screens.
The profile dropdown rendered the wallet item unconditionally, so it
still showed after an admin disabled the personal/topup (wallet) sidebar
module. Reuse the sidebar module visibility check so the dropdown honours
the same toggle as the sidebar.
Fixes#5696
* fix(markdown): render default markdown with marked
- switch default frontend markdown rendering from react-markdown/remark-gfm to marked to avoid old WebKit parse failures from lookbehind regex literals
- sanitize marked HTML output with DOMPurify and preserve external link target and rel behavior
- remove default direct dependencies on react-markdown, remark-gfm, and rehype-raw while leaving classic unchanged
* fix(markdown): expand default markdown rendering support
- render default markdown with marked extensions for KaTeX formulas, page breaks, and common emoji shortcodes.
- sanitize KaTeX output with an explicit DOMPurify allowlist while preserving external link behavior.
- avoid overriding marked text rendering so lists and inline parsing keep their internal parser context.
* fix(markdown): render diagram code blocks in default UI
- add sanitized SVG rendering for flow and sequence diagram code blocks.
- size flow nodes from their labels and route edges from node anchors to prevent clipping.
- style diagram nodes, arrows, labels, and notes with theme-aware classes.
* feat(dashboard): add traffic flow sankey chart
Add dashboard flow APIs and a Sankey-based flow view with user, optional API key, model, and channel layers.\n\nReuse the dashboard VChart palette, add precise link/node tooltips and interactions, and cover filtering, layer ordering, color stability, and error states with tests.
* feat: build flow chart from quota data
---------
Co-authored-by: CaIon <i@caion.me>
Add an opt-in card view to the shared data-table stack (DataTablePage),
toggled via a segmented control in the toolbar with per-table localStorage
persistence. Cards render generically from column meta by default, with an
optional renderCard slot. Defaults to table-only so existing pages are
unchanged.
Wire it into the channels page with a bespoke ChannelCard that reuses every
column's cell renderer, preserving all table information and interactions
(selection, inline priority/weight, balance refresh, status, actions, tag
expand).
- apply table-level truncation only to primitive text cells.
- let tables expand to content width with horizontal scrolling to avoid column overlap.
- update model badges to show full model names on one line.
TanStack row objects may keep a stable reference while their selection
state changes, so reading row.getIsSelected() inside the React.memo
comparator could miss updates. Capture isSelected as an explicit prop
and compare it instead, ensuring rows re-render when selection toggles.
* fix(ui): prevent table cell text overflow
- add default truncation with hover details for text cells in shared and static data tables to prevent content from spilling into adjacent columns.
- adjust API key group, model, and IP restriction columns to fix badge overlap and left alignment drift.
- reuse a shared truncated cell component and add width constraints for composite badge cells.
* fix(table): prevent badge content from overflowing columns
- make table text and badge cells shrink within constrained columns so long values truncate instead of bleeding into adjacent cells.
- add a shared BadgeCell wrapper to keep badge alignment consistent across API keys and other list pages.
- update affected list views to use constrained wrappers for group, provider, pricing, OAuth, and API info values.
* perf(table): use percentage-based column widths
- compute each column's width as a percentage of total column size instead of a fixed pixel value, letting the colgroup scale fluidly with the table container.
* perf(data-table): reduce unnecessary re-renders across table components
- stabilize commitSearchValue in toolbar by reading table/searchKey via refs, eliminating recreation on every parent render
- store onColumnFiltersChange in a ref so debounce effect is not reset when the caller passes a new function reference each render
- wrap DataTableRow in React.memo with a custom comparator that ignores getColumnClassName reference churn
- memoize selectedValues Set in DataTableFacetedFilter and wrap with React.memo to prevent rerenders on unrelated state changes
- cache cell meta reads in CompactRow and FallbackRow to a single pass per row; memoize hasCompactMeta in MobileCardList
- memoize hideable columns list in DataTableViewOptions and colSpan in DataTableView
- remove tableClassName and colgroup from scroll-sync effect deps; cache toolbar button NodeList via useLayoutEffect to avoid per-keydown DOM queries
* perf(data-table): replace scroll-sync split header with CSS sticky
- remove JS scroll-sync effect and event listener between split header and body containers.
- merge separate header/body tables into a single scrollable table element, reducing DOM complexity.
- apply CSS sticky positioning to the header for a simpler, hardware-accelerated freeze effect.
* fix(data-table): replace opacity muted colors with color-mix
- switch from bg-muted/50 and bg-muted/30 to color-mix(in oklch) to produce opaque blended backgrounds that prevent scroll content from showing through pinned cells.
- expose --table-header-bg CSS variable so pinned header cells inherit the exact same computed color as the thead background.
- add group class to TableRow to enable group-hover selectors on pinned cell styles.
* feat(data-table): support column pinning via meta.pinned
- add pinned?: 'left' | 'right' to ColumnMeta so pinning is declared once in the column definition and applies to both header and body automatically
- DataTableView derives pinnedColumns from meta.pinned at runtime, merged with any explicit pinnedColumns prop; explicit entries take precedence
- add header and meta.pinned: 'right' to all actions columns across channels, users, api-keys, redemption-codes, models, deployments, and subscriptions tables
* style(row-actions): align action buttons to leading edge of column
* refactor(data-table): extract BadgeListCell and centralize badge alignment
- add BadgeListCell component to data-table for badge lists with overflow tooltip, replacing duplicated renderLimitedItems helpers in channels, models, and pricing columns
- move StatusBadge -ml-1.5 alignment into the component itself via a table-cell context selector, so callers no longer need manual offset wrappers
- remove the table-cell-level -ml-1.5 selector from TableCell now that alignment is handled by StatusBadge directly
* style(row-actions): offset action buttons to align with column header text
* refactor(data-table): consolidate mobile meta into ColumnMeta declaration
- move mobileTitle, mobileBadge, mobileHidden into the global ColumnMeta augmentation so the type is shared across the project
- remove the local MobileColumnMeta interface and getCellMeta helper from mobile-card-list.tsx
- direct col.columnDef.meta access is now type-safe without explicit casting
* refactor(data-table): simplify column header and meta config
- auto-render string `header` values via DataTableColumnHeader so sortable/non-sortable columns work without boilerplate function wrappers
- promote mobile layout hints (mobileTitle, mobileBadge, mobileHidden) into the global ColumnMeta type, removing the local MobileColumnMeta cast in mobile-card-list
- migrate all column files from `meta: { label }` to top-level `header: t('...')`, cutting ~180 lines of repetitive template code
- ViewOptions and MobileCardList label resolution now reads string header first, then meta.label as fallback
* feat(status-badge): add text and underline display types
- introduce StatusBadgeType ('badge' | 'text' | 'underline') and StatusBadgeTypeContext so ancestors can override rendering without touching call sites
- mobile card field rows now use the text type via context, showing badges as plain colored text instead of pills
- ProviderBadge gains data-slot='provider-badge' to enable targeted CSS resets in compact layouts
- replace the implicit [[data-slot=table-cell]>&]:-ml-1.5 rule with explicit -ml-1.5 at each column call site
* refactor(data-table): simplify table filtering internals
- derive toolbar search state from the active table filter to avoid render-time ref writes.
- extract faceted filter selection updates into a pure helper for clearer single and multi-select behavior.
- split pinned column resolution into focused helpers so explicit and meta pins merge predictably.
Co-authored-by: t0ng7u <dev@aiass.cc>
* refactor(web): centralize data table implementation
- route all TanStack table setup through a shared data-table hook to remove repeated state and row model wiring.
- move table rendering, static table wrappers, empty states, and primitive exports behind the data-table module.
- update feature tables and configuration editors to share the same table UX while preserving their existing workflows.
* refactor(web): trim data table public API
- remove unused data-table exports and dead static table helper types.
- keep internal table header, skeleton, empty state, and faceted filter helpers private to the data-table module.
- route feature imports through the data-table barrel to avoid subpath coupling.
* refactor(web): unify table rendering components
- centralize static table headers, bodies, empty states, and shared class names behind the data-table package.
- migrate settings, pricing, channel, key, subscription, and model tables to the shared table APIs.
- remove data-table exports for low-level table primitives so feature code uses one supported abstraction.
* perf(web): keep list tables fixed within page content
- make shared data table pages fill available height and scroll row data inside the table body.
- add a fixed content layout mode so selected list pages avoid page-level scrolling.
- apply the fixed table behavior to keys, logs, channels, models, users, redemptions, and subscriptions.
* perf(web): refine table pagination controls
- show total row counts instead of redundant page range text.
- tighten visible page buttons so pagination fits constrained table widths.
- align pagination controls and tune text hierarchy for clearer scanning.
* perf(web): stabilize model pricing table columns
- keep model pricing columns at fixed widths so headers do not collapse in narrow layouts.
- truncate long model names and pricing summaries within their cells instead of squeezing adjacent columns.
* refactor(web): simplify data table rendering internals
- split table body rendering into focused helpers for loading, empty, and row states.
- extract static table row and cell class resolution to reduce branching in the main component.
- reuse a single pagination page-size option list to avoid duplicated constants.
* perf(pricing): reduce dynamic pricing table render work
- reuse dynamic pricing field metadata instead of rebuilding it inside table columns.
- precompute formatted dynamic prices per tier and group to avoid repeated entry mapping for each cell.
- simplify select option construction in related dialogs while preserving the same choices.
* refactor(web): streamline pricing table rendering
- reuse translated endpoint select options between trigger data and menu items.
- precompute dynamic pricing maps per group so table cells only resolve formatted values.
- add local dynamic pricing type aliases to keep helper signatures readable.
* refactor(web): merge pricing table imports
* refactor(web): merge upstream ratio table imports
* refactor(web): merge channel selector table imports
* refactor(web): simplify tiered pricing select items
* refactor(web): reuse model ratio row state
* refactor(web): rely on table view row defaults
* refactor(web): reuse pagination state values
* refactor(web): hoist pagination size select items
* refactor(web): clarify static table body rows
* refactor(web): extract table page pagination rendering
* fix(web): remove direct hast type dependency
- rely on Shiki transformer contextual typing for line nodes.
- allow frontend typecheck to pass without an undeclared hast package.
* refactor(web): trim data table hook return API
- return only the TanStack table instance from useDataTable.
- keep internal state handling private because callers do not consume it directly.
* refactor(web): keep static table empty row private
- stop exporting the internal StaticDataTableEmptyRow helper.
- keep the public static table API focused on the table component and column type.
* refactor(web): hide data table view props from barrel
* refactor(web): remove stale long text lint override
* fix(web): keep pinned table columns opaque
- apply pinned column background classes after custom column classes.
- use an opaque hover background so scrolled content cannot show through fixed cells.
* refactor(data-table): organize shared table components
- group table primitives, page composition, toolbar controls, static tables, and hooks by responsibility.
- split shared view types, row rendering, header rendering, and pinned-column styling out of the main table view.
- keep the public data-table barrel stable while documenting the new ownership boundaries.
* fix(web): stabilize split table column sizing
- derive default colgroup widths from visible columns when split headers or header sizing are enabled.
- apply a fixed table layout with computed minimum width so header and body columns stay aligned.
- keep split-header containers from leaking horizontal overflow and avoid extra pinned-column borders.
* fix(web): set stable table utility column widths
- assign fixed widths to selection columns so shared colgroup sizing keeps checkbox cells compact.
- size id columns in redemption and user tables to keep split headers aligned with body rows.
* fix(web): align model metadata icon cells
- render compact provider avatars in the metadata icon column instead of wide wordmarks.
- position icons in a fixed-size wrapper so they line up with the existing icon header alignment.
* fix(status-badge): hide status dot by default
* fix(web): prevent user invite info overlap
- give the invite info and created-at columns explicit widths so table sizing reserves enough space.
- allow invite badges to wrap within the cell instead of spilling into adjacent columns.
* perf(data-table): cache pinned column class resolution
- reuse the pinned column lookup while table props stay stable to reduce repeated per-render work.
- share the resolved column class handler across unified and split-header table layouts.
- localize page-number screen reader labels so pagination remains accessible in every locale.
* refactor(data-table): tighten static table modes
- make StaticDataTable distinguish data-driven and children-only usage through explicit prop shapes.
- remove unsupported columns-without-data fallback after confirming no repository callers rely on it.
- default manual table modes away from unused local row models to reduce repeated table work.
* fix(data-table): make pinned edit column opaque
- use an opaque muted background for the active action column so sticky cells do not reveal scrolled content underneath.
* fix(data-table): prevent narrow column overlap
- apply stable header sizing to remaining desktop data table pages so constrained layouts scroll instead of compressing cells.
- add explicit widths for key, quota, badge, and timestamp columns that contain fixed-format content.
- constrain masked values and timestamp cells with truncation to keep content inside its assigned column.
* fix(table): align table cell content with headers
- remove extra inline padding from masked table text buttons so values start at the cell edge.
- tag status badges and offset leading badges inside table cells to match header text alignment.
* fix(table): prevent admin list column overflow
- widen redemption and subscription table columns so masked codes, timestamps, and localized headers fit.
- localize subscription ID headers and add Received amount translations across supported locales.
* fix(provider-badge): unify provider icon spacing
- add a shared provider badge component for icon and status label layout.
- reuse it in channel type and model vendor columns so OpenAI icons align consistently.
- migrate frontend dialogs to the shared footer API so actions stay separated from scrollable body content.
- tune dialog dimensions for model analytics, prefill groups, billing history, channel model sync, and related workflows.
- update channel terminology and dialog action translations across supported locales.
- introduce a reusable dialog component for consistent header, body, and footer layout.
- support per-dialog sizing, trigger rendering, initial focus, and close button controls.
- preserve base dialog open and close motion classes while allowing content-specific styling.
- introduce a shared themed JSON editor with line numbers, formatting, status feedback, and keyboard editing helpers.
- use the shared editor in model pricing JSON mode so pricing maps get consistent editor behavior.
- localize structured JSON validation messages so parse errors avoid browser-specific English text.
- scope validation queries with a form root id so feedback stays inside the submitted form.
- scroll to the earliest invalid control or message and move focus without fighting scroll position.
- avoid handling the same failed submit twice to keep retries from jumping unexpectedly.
Implement the Simple Large-font theme preset and xl font scale options to enhance interface accessibility. Remove status indicator dots from channel badges in logs to keep the table layout visual and clean.
System settings forms that used flat dotted API keys (e.g.
`performance_setting.monitor_cpu_threshold`) with React Hook Form were
broken: RHF stores dotted paths as nested objects on update, while dirty
checks and submit comparisons still read flat keys from defaults. Users
could edit values but always saw "No changes to save".
Refactor affected sections to use nested Zod schemas and default values
for RHF, with explicit helpers to convert between nested form state and
flat API keys. Track a normalized baseline in refs for accurate change
detection and post-save resets.
Add `safeNumberFieldProps` to prevent native `<input type="number">`
from writing NaN into form state when cleared. NaN caused Zod validation
to fail silently and made the save button appear unresponsive. The
helper ignores non-finite updates so controlled inputs snap back to the
last valid value, matching legacy Semi InputNumber behavior.
Sections refactored for dotted-key handling:
- maintenance/performance-section
- models/grok-settings-card
- auth/passkey-section
- auth/oauth-section
- auth/section-registry (pass attachment_preference raw; normalize in section)
Sections migrated to safeNumberFieldProps:
- maintenance/performance-section
- models/grok-settings-card
- integrations/monitoring-settings-section
- integrations/payment-settings-section
- integrations/creem-product-dialog
- general/pricing-section (USD exchange rate)
- general/system-behavior-section
- content/dashboard-section
Optional numeric fields (e.g. custom currency exchange rate) keep their
existing empty-to-undefined semantics and are intentionally unchanged.
- Restore StatusBadge sizes to h-5/text-xs (sm, md) and h-6/text-xs (lg).
- Restore classic textColorMap coloring and status indicator dot.
- Embed channel type icon directly inside StatusBadge as custom children.
- Re-align status badge colors: danger for manual disabled, warning for auto disabled.
Introduce a switchable Anthropic-inspired color preset and a new Font customization axis so users can adopt the editorial serif look across the entire UI, including sidebar navigation, tabs, form controls, buttons, and table headers.
Theme preset
Add anthropic to the theme preset registry with warm cream canvas, slate foreground, and clay/coral accent tokens for light and dark modes
Define explicit surface colors for the Anthropic preset instead of relying on the semantic surface bridge
Exclude anthropic from the primary-color surface bridge so bespoke warm neutrals are not overridden by accent-tinted mixes
Typography system
Add @fontsource-variable/lora and a global --font-serif token with CJK serif fallbacks (Noto Serif SC, Source Han Serif, Songti SC, etc.)
Introduce a --font-body token and drive <body> font-family from it
Add a Font axis (default | sans | serif) parallel to radius/scale
Resolve font: 'default' against preset defaults (anthropic → serif)
Persist font preference via cookie and apply data-theme-font on <body>
Apply serif OpenType features (kern, liga, calt, tnum) and heading display tuning when serif is active
Remove per-component sans opt-outs so serif inherits through sidebar, tabs, inputs, buttons, badges, and table headers via natural CSS cascade
Keep monospace contexts unchanged via Tailwind preflight and .font-mono
UI and i18n
Add Font selector to the theme config drawer (Auto / Sans / Serif)
Add "Font" and "Select body font" translations for en, zh, fr, ja, ru, vi
Misc
Tighten group and status badge sizing for better balance with serif text
Restructure the default-theme channel create/edit experience to match classic
frontend behavior, improve form UX, and align with the project's Base UI design
system.
Channel editor architecture:
- Split the monolithic channel mutate drawer into focused section components
(basic, API access, auth, models, advanced) with shared drawer layout
primitives
- Extract submission, toast handling, and react-query cache invalidation into
`useChannelMutateForm`
- Add a dedicated loading skeleton for channel detail fetch during edit mode
- Remove the top-level configuration summary block
Form validation and data handling:
- Strengthen `channel-form` Zod schema with JSON, model mapping, status code
mapping, Codex credential, and Vertex AI key refinements
- Move type-specific conditional validation into `superRefine`
- Normalize base URL formatting and tighten model mapping value validation
Model mapping editor:
- Add Visual/JSON tabbed editing with inline JSON and duplicate-key feedback
- Improve accessibility for icon-only actions and add model suggestion datalists
MultiSelect component:
- Replace the custom cmdk-based implementation with Base UI Combobox chips
- Align focus, border, ring, disabled, and invalid states with standard Input
styling via `ComboboxChips`
- Preserve existing API for all current callers (`options`, `selected`,
`onChange`, `allowCreate`, `createLabel`)
- Support inline custom value creation and comma/newline batch input
- Limit visible chips with a compact "+N more" overflow summary via
`maxVisibleChips` (8 in the channel editor)
- Anchor the dropdown to the full chips container via `useComboboxAnchor` so
the popup matches input width and long model names are no longer truncated
Models & groups UX:
- Integrate manual custom model entry directly into the model MultiSelect
- Remove the separate manual model input/button block
- Keep selected-model count badge and existing model-mapping guardrail behavior
i18n:
- Add and sync translation keys for section descriptions, validation messages,
model mapping UI, and MultiSelect labels across en, zh, fr, ja, ru, and vi
- Fix missing translations for "Name, provider type, and availability.",
"Endpoint, provider-specific settings, and credentials.", and "Published
models, groups, and model remapping rules."
- Remove obsolete keys tied to the deprecated summary and manual model entry UI
Restructure the default-theme channel create/edit experience to align with
classic frontend behavior, modern form UX patterns, and the project's Base UI
design system.
Channel editor architecture:
- Split the monolithic channel mutate drawer into focused section components
(basic, API access, auth, models, advanced) with shared drawer layout
primitives
- Extract submission, toast handling, and react-query cache invalidation into
`useChannelMutateForm`
- Add a dedicated loading skeleton for channel detail fetch during edit mode
- Remove the top-level configuration summary block per UX feedback
Form validation and data handling:
- Strengthen `channel-form` Zod schema with JSON, model mapping, status code
mapping, Codex credential, and Vertex AI key refinements
- Move type-specific conditional validation into `superRefine`
- Normalize base URL formatting and tighten model mapping value validation
Model mapping editor:
- Add Visual/JSON tabbed editing with inline JSON and duplicate-key feedback
- Improve accessibility for icon-only actions and add model suggestion datalists
MultiSelect component:
- Replace the custom cmdk-based implementation with Base UI Combobox chips
- Align focus, border, ring, disabled, and invalid states with standard Input
styling via `ComboboxChips`
- Preserve existing API (`options`, `selected`, `onChange`, `allowCreate`,
`createLabel`) for all current callers
- Support inline custom value creation, comma/newline batch input, searchable
options, portal-based dropdown positioning, and chip removal
Models & groups UX:
- Integrate manual custom model entry directly into the model MultiSelect
- Remove the separate manual model input/button block
- Keep selected-model count and existing model-mapping guardrail behavior
i18n:
- Add and sync translation keys for new editor sections, validation messages,
model mapping UI, and MultiSelect empty/create labels across en, zh, fr, ja,
ru, and vi
- Remove obsolete keys tied to the deprecated summary and manual model entry UI
Affected areas:
- `web/default/src/features/channels/components/drawers/`
- `web/default/src/features/channels/hooks/use-channel-mutate-form.ts`
- `web/default/src/features/channels/lib/channel-form.ts`
- `web/default/src/features/channels/lib/model-mapping-validation.ts`
- `web/default/src/features/channels/components/model-mapping-editor.tsx`
- `web/default/src/components/multi-select.tsx`
- `web/default/src/i18n/locales/*.json`
Refactor the usage log filter toolbar into a shared reusable component for common, drawing, and task logs. Optimize desktop filters with a responsive grid, move secondary filters into a mobile drawer, standardize filter typography, remove redundant filter icons, and add the missing i18n translations for the new drawer description.
Resolve verified V1 frontend feedback by improving channel workflows, auth behavior, API key interactions, user filtering, layout persistence, subscription quota handling, i18n text, pricing metadata, and stale frontend cache recovery.
- Add a global frontend cache version cleanup to prevent old frontend localStorage from causing page errors after upgrades.
- Fix channel copy refresh, model mapping input focus loss, create-channel fetch-model title state, upstream model update confirmation, and batch test toast behavior.
- Respect password login settings and improve Turnstile, forgot-password, registration, and invite-link flows.
- Make user role/status filtering server-side and preserve table page size in URLs.
- Improve API key edit validation feedback and prefetch real keys for reliable copy actions.
- Fix rankings access fail-open behavior, double scrollbars, subscription received amount conversion/display, token i18n wording, model deletion confirmation grammar, and Claude pricing context inference.
- Add clearer Playground model/group loading errors.
Validation:
- bun run typecheck
- bun run i18n:sync
- gofmt on modified Go files
- go test ./controller ./model -run '^$'
QuentinHsu