Follow-up to #23573: chip-heavy answers are long by design, and each
stream flush re-ran `protectChatReferencesForMarkdown` and
`marked.lexer` over the whole message, so render cost grew quadratically
with message length. This makes the per-flush cost proportional to the
appended text instead, and offsets the new code by removing dead AI chat
code.
## Streaming render
- **Incremental block splitting.** `getMarkdownBlocksIncrementally`
reuses blocks that can no longer change and re-tokenizes only the
trailing ones. Two trailing blocks stay unstable, not one: a loose list
followed by a blank line still merges with a later item (`- a\n\n` + `-
b` is one list token). Uses `Lexer.blockTokens` instead of
`marked.lexer` since only block raws are needed and the full lexer also
runs the inline tokenizer. Simulated stream over a 22 KB chip-heavy
message (120 chars/flush, matching the 100 ms flush throttle): 191 ms →
3.7 ms cumulative. The test suite pins char-by-char equivalence against
full `marked.lexer` output across loose lists, unclosed fences, setext
headings, tables, CRLF and chip markers.
- **Per-block reference protection.** `protectChatReferencesForMarkdown`
moved behind the existing block memo, so settled blocks never re-run
reference parsing during a stream.
- **Anchored open pattern.** `(?<!\[)\[\[+` anchors marker matching to
the start of a bracket run. The greedy `+` from #23798 backtracked at
every position inside a run, once per alternative: 429 ms → ~1 ms on a
10 KB bracket-run input. A run start always yields the same match, so no
valid marker is lost. Also an `includes('[[')` bail-out in
`findChatReferences`, which runs on every text node of the streaming
block.
## Chip lookups
`fieldMetadataItemByIdSelector` did `objectMetadataItems.find(obj =>
obj.fields.some(...))` per chip — O(workspace fields) each time the
agent's tool calls trigger a metadata refetch mid-chat. The by-id and
by-name map selectors mostly already existed with almost no consumers;
this wires `fieldMetadataItemByIdSelector`,
`objectMetadataItemFamilySelector` and `viewFromViewIdFamilySelector` to
them (adding the missing `objectMetadataItemsByIdMapSelector` and
`viewsByIdMapSelector`) and adds `areEqual` so unchanged lookups keep
referential stability.
## Offscreen messages
Settled messages (everything except the streaming last one) get
`content-visibility: auto`, so long threads skip layout and paint for
messages scrolled out of view. `contain-intrinsic-size: auto` keeps
remembered heights, so scroll positions stay accurate once a message has
been painted.
## Removed
`ReasoningSummaryDisplay`, `agentChatMessagesComponentState`,
`CHAT_THREADS_PAGE_SIZE`, `AgentResponseFormat` and `getFieldIcon` had
no consumers. `TextWithChatReferences` and
`protectChatReferencesForMarkdown` shared a duplicated segment-slicing
loop, now in `getChatReferenceSegments`, and the nine identical per-tag
markdown component entries collapse into `createChatReferenceElement`.
The branch lands at +354/−329 including the new test suite; production
code is net negative.
Incidental: `marked` added to jest's `transformIgnorePatterns` allowlist
(ESM-only, previously imported by no test).
---
_Generated by [Claude
Code](https://claude.ai/code/session_01MN8FVc63J4SJQHXWwzUwkh)_
<!-- This is an auto-generated description by cubic. -->
<a
href="https://cubic.dev/pr/twentyhq/twenty/pull/23831?utm_source=github"
target="_blank" rel="noopener noreferrer"
data-no-image-dialog="true"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://www.cubic.dev/buttons/review-in-cubic-light.svg"><img
alt="Review in cubic"
src="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"></picture></a>
<!-- End of auto-generated description by cubic. -->
Follow-ups recorded in #23023, done in one pass.
## Shared primitives
- Folded `PageLayoutWidgetSortableItem` and `PageLayoutWidgetDropLine`
into the shared `DragDropItemSortableCell` / new `DragDropItemDropLine`
(new `data`, `dropLine`, `highlightWhileDragging`, `hasTransition`
props).
- Added generic `DragDropProviderDragStartEvent` (and
DragMove/DragOver/DragEnd/DropTarget) helpers and deleted the 7 copied
`Parameters<...>` extractions across the dnd hooks.
- Replaced the `useMovePageLayoutWidgetUp/Down` implementations (~140
lines) with `moveWidgetWithinTabInDraft`.
- Migrated the remaining page-layout test suites onto
`pageLayoutDraftFixtures`.
## Tab reordering off Pangea
- Tabs are sortable cells on the same provider as widget drags,
segregated by dnd type, so widget drops on tab buttons keep working
while tabs reorder.
- Reordering is ID based (`reorderTabInDraft`: insert before the hovered
tab), which keeps the pinned first tab in place without index
arithmetic.
- Preserved overflow behaviors: the dropdown stays open while a tab drag
is in flight, dropping a tab on the "+N More" button appends it and
opens the dropdown, and both the visible strip and the overflow list
have end drop zones.
## Fields configuration editors off Pangea
- Group reorder, field reorder and cross-group field moves now run on
the shared cells (same drop line and end-zone patterns).
## DraggableList off Pangea
- `DraggableList` / `DraggableItem` keep their consumer-facing API — the
~9 consumers now type their handlers with a local
`DraggableListDropResult` instead of pangea's `DropResult` — but run on
the shared sortable cells; each list's uuid group doubles as its dnd
type so nested lists stay isolated from page-level providers.
- Items register their index in a list-scoped registry so the end drop
zone can resolve the append index at drop time (with insert-before
semantics an item could otherwise never reach the last position).
- Deleted three dead files that only existed for pangea plumbing (the
side panel navigation placeholder, `getCssCompatibleDraggableProps`, the
orphaned `recordGroupPendingDragEndReorderState`).
## Record table row drag off Pangea
- Rows register through `useSortable` directly on the row element — no
wrapper div, so row CSS, sticky cells and virtualization stay untouched
— with the grip cell wired as the drag handle via the shared sortable
handle ref context.
- Both table modes (virtualized flat list and record groups) share a
`DragOverlay` clone that replaces pangea's virtual-mode `renderClone`,
and end drop zones per record group (and after the virtualized list)
allow dropping after the last row or into an empty group.
- The drop handlers keep their pangea-shaped result object, retyped as a
local `RecordDragDropResult`, so the position computation logic is
untouched.
## Pangea removed
`@hello-pangea/dnd` is gone from `package.json` and the lockfile, along
with its orphaned transitive entries (`css-box-model`, `raf-schd`,
`react-redux`, `redux`). Nothing in the repo imports it anymore.
## Dashboards: cross-tab widget drag for grids
react-grid-layout drags never enter dnd-kit, so the bridge hit-tests the
pointer against the tab buttons' `data-page-layout-tab-drop-target-id`
rects during grid drags, highlights the hovered tab through state, and
on drop moves the widget to the destination grid below its existing
content (`moveWidgetToGridTabInDraft`, `buildTabWidgetLayouts`). The
grid's own post-drag layout commit is suppressed once so it does not
overwrite the cross-tab move.
## Fixes found while testing
- With `feedback: 'clone'`, the drag source is its own initial drop
target and its placeholder is a DOM clone taken at drag start, so the
drop line rendered into the source got baked into the placeholder and
stuck there for the whole drag. The line is now hidden on the source
cell, leaving a single indicator at the actual target.
- Reorderable tabs collapsed to text height and sat top-aligned next to
"+ New Tab" because the sortable cell wrapper defaults to `display:
block; height: auto`, breaking the tab height chain — the tab list now
uses the cell's `fill` mode so tabs stretch to the strip height again.
## Testing
Playwright against the dev app:
- Record page: widget reorder up and down in the pinned column (single
blue drop line at the target), drag to another tab via its tab button
(highlight + move), drag back into content at a specific position,
chained cross-tab moves, tab reorder with vertical drop line, new tab
creation.
- Overflow (narrow viewport): drop a tab on "+N More" (appends last,
dropdown opens), reorder inside the dropdown (stays open), drag a tab
from the dropdown back to the visible strip.
- Dashboard: grid drag within a tab, cross-tab drag onto a tab button
(hover highlight, widget lands below destination content, remaining
widgets keep their positions), save and reload persistence in both
directions.
- Fields editor: field reorder, group reorder, field move across groups,
plus the Move Up / Move Down widget actions.
Since the pangea-removal commits:
- Typecheck, oxlint and oxfmt green over the full front source; unit
suites green including the migrated `useStartRecordDrag` test (jest
needed a scoped transform exemption for `@preact/signals-core` once
dnd-kit reached the side-panel suites).
- Storybook visual regression unchanged across ~700 stories — expected,
since the migrated surfaces render identical DOM at rest (drop lines and
drag overlays only exist mid-drag).
- The tab strip fix reverses the exact regression mechanism: the
sortable cell wrapper defaulted to `display: block; height: auto`,
collapsing the tab height chain next to the full-height "+ New Tab"
button; `fill` restores the stretch.
---
_Generated by [Claude
Code](https://claude.ai/code/session_01XKRCzzu8oGyocXZtFp7VEG)_
<a
href="https://cubic.dev/pr/twentyhq/twenty/pull/23211?utm_source=github"
rel="nofollow noreferrer noopener" target="_blank">``<img alt="Review
in cubic"
src="https://www.cubic.dev/buttons/review-in-cubic-dark.svg">``</a>
## Desktop
https://github.com/user-attachments/assets/87999ff5-9257-4fed-90e0-781322b1db98
## Mobile
https://github.com/user-attachments/assets/0d115cb6-7bca-4229-8fde-0f188e862b00
Greets the user the moment they finish onboarding: a halftone Twenty
mark (brand purple) assembles from scattered particles, shimmers, then
bursts outward to reveal the freshly loaded workspace behind it. Plays
once, ~3s, with a reduced-motion fallback.
`PageChangeEffect` sets a transient `isWelcomeAnimationVisibleState`
atom at the onboarding to workspace redirect seam (gated by
`shouldShowWelcomeAnimationOnNavigate`); `WelcomeOverlay`, mounted in
`WorkspaceAppProviders`, plays it and clears it. Transient by design, so
it never replays on reload.
## Why a canvas rendered on another thread
The overlay plays at the exact moment the workspace mounts behind it,
which is one of the heaviest main-thread stretches in the app (metadata,
providers, first data fetch). Anything animating on the main thread
competes with that work:
- framer-motion and plain main-thread `requestAnimationFrame` visibly
stuttered and froze during the mount, because long tasks starve rAF.
- CSS keyframes stay smooth (they run on the compositor), but can't
drive a ~2900-particle halftone with per-particle physics (staggered
assemble, moving shimmer band, directional burst).
So the halftone runs as a Canvas 2D particle system inside a Web Worker,
drawing to an `OffscreenCanvas` transferred from the main thread. The
whole render loop lives off the main thread, so app-mount jank can't
reach it. A main-thread path is kept as a fallback for browsers without
`OffscreenCanvas` / `transferControlToOffscreen`, and Safari workers
(which lack rAF) fall back to a fixed-interval loop. The renderer is a
pure, DOM-free module so it bundles cleanly into the worker; dot colors
come from CSS vars read on the main thread and passed in.
Also fires for every onboarding completion path and skips the billing
`PlanRequired` detour.
<!-- This is an auto-generated description by cubic. -->
<a
href="https://cubic.dev/pr/twentyhq/twenty/pull/22622?utm_source=github"
target="_blank" rel="noopener noreferrer"
data-no-image-dialog="true"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://www.cubic.dev/buttons/review-in-cubic-light.svg"><img
alt="Review in cubic"
src="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"></picture></a>
<!-- End of auto-generated description by cubic. -->
## What
Adds a custom Jest reporter to **twenty-front** that, in CI, suppresses
passing-test output and surfaces only failures — mirroring what we
already do for **twenty-server**.
## How
- New `packages/twenty-front/jest-failures-only-reporter.cjs` — a
verbatim port of
`packages/twenty-server/jest-failures-only-reporter.js`. It prints a
`FAIL` block per failing suite plus a final "FAILED TEST SUITES
SUMMARY", and otherwise emits only the suite/test totals.
- Wired into `packages/twenty-front/jest.config.mjs` via `...(isCI && {
reporters: ['./jest-failures-only-reporter.cjs'] })`, gated on `CI ===
'true'` exactly like twenty-server.
### Note on the `.cjs` extension
twenty-front's `package.json` sets `"type": "module"`, so a `.js`
reporter is parsed as ESM and `module.exports` throws. Renaming to
`.cjs` keeps the file as CommonJS (Jest requires CJS reporters).
twenty-server is not an ESM package, hence its `.js` extension.
## Testing
- Passing suite (`CI=true npx jest <file>`): output reduced to the
totals summary only.
- Temporary failing suite: shows the `FAIL` block, failure message, and
the failed-suites summary.
Local dev runs (no `CI` env) are unaffected — the default reporter is
used.
<!-- This is an auto-generated description by cubic. -->
<a
href="https://cubic.dev/pr/twentyhq/twenty/pull/22345?utm_source=github"
target="_blank" rel="noopener noreferrer"
data-no-image-dialog="true"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://www.cubic.dev/buttons/review-in-cubic-light.svg"><img
alt="Review in cubic"
src="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"></picture></a>
<!-- End of auto-generated description by cubic. -->
Migrates `twenty-front`, `twenty-sdk`, and
`twenty-front-component-renderer` from `twenty-ui-deprecated` to
`twenty-ui` (mechanical import swap — the packages have API parity) and
deletes the deprecated package along with its workspace/CI/config
wiring.
Also adds `@linaria/react`/`@linaria/core` as direct deps of
`twenty-front` (it used them transitively via the deprecated package).
Note: move the required status check from `ci-ui-status-check` to
`ci-new-ui-status-check`.
Argos: the Storybook box-model/button-reset baseline shift (the bulk of
the visual diffs) is isolated in #21665 — Storybook now loads
twenty-ui's global `reset.scss`, which the production app already ships.
Once #21665 merges and this branch is rebased, the remaining Argos diffs
are component-level visual-parity items only.
## Description
Promotes the next-gen UI library (formerly `twenty-new-ui`) to the name
**`twenty-ui`** (v0.1.0, publishable) and renames the old package to
**`twenty-ui-deprecated`**. Rewrites ~1,730 `twenty-ui` imports →
`twenty-ui-deprecated`, updates all configs/CI/Docker/deps, and migrates
twenty-front's `Toggle` to the new package (first consumer) as a
drop-in.
## Next steps
- Wire the `ui/v*` publish dispatch (`cd-deploy-tag.yaml` +
`.yarnrc.yml`), then tag `ui/v0.1.0` to publish.
- Continue migrating components from `twenty-ui-deprecated` →
`twenty-ui`.
## Summary
Iterative redesign of two related areas in settings, plus a new
`pages/settings/layout/` folder for read-only entity detail pages.
### Application content tab
- **Grouped into three sections** — Data / Layout / Logic — each with
one H2 + multiple `TableSection`-wrapped sub-tables (mirrors the
role-permissions pattern). Replaces six per-category table/row
components with one uniform `<SettingsApplicationContentSubtable>` +
`ApplicationContentRow` shape (net **−~700 lines** across the refactor).
- **All 10 row categories now clickable** for installed apps:
- Objects / Fields / Logic functions / Front components → existing
detail pages
- Agents → existing `AiAgentDetail`
- Skills → existing `AiSkillDetail` (looked up by `Skill.applicationId +
name`)
- Roles → existing `RoleDetail` (looked up by
`Role.universalIdentifier`)
- Views / Page layouts / Navigation menu items → **new** detail pages
(see below)
- **Lifecycle hooks visible** — `pre-install` / `post-install` logic
functions are surfaced in the Trigger column instead of appearing as
empty/misconfigured.
### Logic function settings (Triggers + Test tabs)
- Triggers tab is now editable (HTTP / Cron / Database event / AI tool)
with a `<SettingsLogicFunctionTriggerSection>` wrapper that owns the
toggle, header, and read-only short-circuit.
- HTTP section gets a Live URL field with copy-to-clipboard.
- Each section shows a **Sample input** preview (the JSON the function
will receive) using the same payload builders the Test tab uses.
- Test tab: **Simulate trigger** buttons that prefill the JSON input
from the configured trigger's schema. Replaces an unclickable `<Select>`
(which auto-disables when there's only one option — the typical case).
- Read-only behavior for installed-app functions: explicit `<Callout>`
notice when there's no trigger; trigger sections render as disabled
controls when there is one.
- Removed the empty Environment Variables section from the Settings tab
(it just told the user to go elsewhere).
### New `pages/settings/layout/` folder
Three new app-scoped detail pages so users can drill into entities the
GraphQL `Application` type doesn't expose by id (keyed by manifest
`universalIdentifier`):
- `ApplicationViewDetail` — type, object, visibility + Fields / Filters
/ Sorts subsections (field UIDs resolved to readable labels via
`useFieldLabelByUid`)
- `ApplicationPageLayoutDetail` — type, object + per-tab subsections
listing widgets
- `ApplicationNavigationMenuItemDetail` — type, destination (resolved),
icon, color, position
Each page reads from the marketplace manifest the parent app page
already loads (no extra queries). Folder set up so a future "Layout"
settings tab can grow here (analogous to the existing `data-model/`
folder under the Data tab).
### Other consistency fixes
- Breadcrumbs on every app-scoped entity detail page now include a
category crumb so users know what they're looking at: `Workspace /
Applications / Timely / Navigation menu items / Time entry`.
- Title fallback for nav menu items uses the resolved destination
(`"Time entry"`) instead of the raw enum (`"OBJECT"`).
- New shared utils: `getNavigationMenuItemDestination`,
`resolveManifestObjectLabel`, `getLogicFunctionTriggerLabel`,
`<MonoText>`.
## Backend changes
Only one minor schema-shape change (additive): added `applicationId` to
the `SkillFields` GraphQL fragment and `universalIdentifier` to the
`RoleFragment` so the new lookups have what they need. Generated
metadata schema patched in-tree to match — regenerate with `nx run
twenty-front:graphql:generate --configuration=metadata` if it drifts.
## Test plan
- [ ] Application content tab on an installed app shows the 3 grouped
sections; rows in each section are clickable
- [ ] Click an Object → existing object detail page
- [ ] Click a Field → existing field-edit page
- [ ] Click an Agent / Skill / Role → existing detail page
- [ ] Click a View / Page layout / Navigation menu item → new read-only
detail page; subsections (Fields/Filters/Sorts for views, per-tab
widgets for page layouts) populate correctly
- [ ] Breadcrumbs on every entity detail page have 5 crumbs ending in
`<Category> / <Entity name>`
- [ ] Logic function Triggers tab: toggle each trigger type on/off, see
the Sample input preview update; for installed apps, sections render as
read-only
- [ ] Test tab: each "Simulate trigger" button prefills the JSON editor
with the matching payload shape
- [ ] Functions list: a function configured as `post-install` shows
"Post-install" in the Trigger column
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-authored-by: claude[bot] <claude[bot]@users.noreply.github.com>
Closes#12847
Implements a two-mode focus management pattern for `SelectableList`
components with search inputs, resolving the conflict between text input
cursor movement and grid/list navigation.
### How it works
**Input mode** (search input focused):
- Left/right arrow keys move the text cursor normally
(`enableOnFormTags: false` on ArrowLeft/ArrowRight hotkeys)
- Up/down arrow keys blur the input and transfer focus to the grid,
entering grid mode
**Grid mode** (search input blurred):
- All arrow keys navigate the selectable list grid
- Pressing up arrow from the top row clears the grid selection and
refocuses the search input, returning to input mode
- Typing any printable character refocuses the search input (wildcard
hotkey with `enableOnFormTags: false`)
### Demo
https://github.com/user-attachments/assets/825ad603-a5f8-4863-8269-3ecf35965847https://github.com/user-attachments/assets/9d07346d-18a0-40fa-8874-21040c11f03d
## PR description
- The command menu in the side panel now reads directly from
`MAIN_CONTEXT_STORE_INSTANCE_ID` instead of snapshotting the main
context store into a separate side-panel instance when opening. This
keeps the command menu always in sync with the current page state
(selection, filters, view, etc.).
- Removed the broadening/reset-to-selection feature (Backspace to clear
context, "Reset to" button) since the command menu no longer maintains
its own copy of the context.
## Video QA
https://github.com/user-attachments/assets/5d5bc664-b6d4-431d-a271-6ce23d8a4ae0
## Summary
Replace the single `metadataVersion` integer with per-entity-type
**collection hashes** for granular metadata staleness detection. The
backend already generates a UUID per flat entity map on each cache
recompute (`crypto.randomUUID()` in `WorkspaceCacheService`); we now
expose these via the minimal metadata endpoint and SSE events so the
frontend can compare and know exactly which entity types are stale.
### Key changes
**Backend:**
- `WorkspaceCacheService.getCacheHashes()` — new public method that
reads only `:hash` keys from Redis without fetching full data
- `MinimalMetadataDTO` — added `collectionHashes: Record<string,
string>` (JSON scalar mapping `AllMetadataName` → collection hash),
removed `metadataVersion`
- `MetadataEventDTO` — added optional `updatedCollectionHash` field to
SSE events
- `MetadataEventsToDbListener` — reads the collection hash for the
affected entity type after cache invalidation and attaches it to the SSE
event before publishing
- `MinimalMetadataService` — no longer queries the workspace table; uses
`getCacheHashes()` for all flat entity maps and maps cache keys to
`AllMetadataName` locally
**Frontend:**
- `metadataCollectionHashesState` — new Jotai atom with
`atomWithStorage` + `getOnInit: true` storing
`Partial<Record<MetadataEntityKey, string>>`
- `mapAllMetadataNameToEntityKey()` — explicit mapping from backend
`AllMetadataName` to frontend `MetadataEntityKey` (23 entries)
- `useLoadMinimalMetadata` — stores `collectionHashes` from server,
computes `staleEntityKeys` by comparing local vs server hashes
- `patchMetadataStoreFromSSEEvent()` — accepts optional
`updatedCollectionHash` and updates `metadataCollectionHashesState`
- All 11 SSE effect components — pass
`eventDetail.updatedCollectionHash` through to the patch function
- `useStaleMetadataEntities` — new hook returning entity keys missing
from collection hashes (not yet loaded/synced)
- `resetMetadataStore()` — also clears collection hashes
- Deleted `metadataVersionState` (superseded by collection hashes)
### Design decisions
- **No change to hash generation** — existing `crypto.randomUUID()` is
sufficient. Hashes are persisted in Redis, survive server restarts, and
change only on `invalidateAndRecompute`.
- **"Collection hash" naming** — used consistently to clarify the hash
represents an entire entity collection (e.g., all views), not a single
record.
- **Mapping localized** — backend `WorkspaceCacheKeyName` →
`AllMetadataName` mapping lives in the minimal metadata service.
Frontend `AllMetadataName` → `MetadataEntityKey` mapping lives in a
local utility. Nothing in `twenty-shared`.
- **Backward compatible** — `collectionHashes` is additive;
`updatedCollectionHash` is nullable.
## Summary
Uniformizes the metadata store to support **all** backend flat metadata
types, introduces a **minimal metadata endpoint** for fast initial
renders, replaces custom localStorage persistence with **Jotai's
built-in `atomWithStorage`**, and wires up a
**MinimalMetadataLoadEffect** for stale-while-revalidate loading.
### Key changes
- **All flat metadata types**: Added `FlatCommandMenuItem`,
`FlatFrontComponent`, `FlatWebhook`, `FlatRole`, `FlatRoleTarget`,
`FlatAgent`, `FlatSkill`, `FlatRowLevelPermissionPredicate`,
`FlatRowLevelPermissionPredicateGroup` — every entity in the backend
`MetadataEntityTypeMap` now has a corresponding frontend flat type
registered in `ALL_METADATA_ENTITY_KEYS` and `MetadataEntityTypeMap`.
- **Minimal metadata endpoint** (`minimalMetadata` GraphQL query): New
backend module (`MinimalMetadataModule`) returns lightweight object
metadata (names, icons, labels, flags) and basic views (id, type, key,
objectMetadataId) plus a `metadataVersion`. This enables fast first
paint before full metadata loads.
- **Jotai `atomWithStorage` for persistence**: Replaced the custom
`MetadataLocalStorageEffect` with Jotai's built-in `atomWithStorage` on
both `metadataStoreState` (family) and `metadataVersionState`. Added
`localStorageOptions` support to `createAtomFamilyState` for `{
getOnInit: true }` synchronous hydration. Each entity atom auto-persists
under keys like `metadataStoreState__objectMetadataItems`.
- **MinimalMetadataLoadEffect**: New effect mounted before
`MetadataProviderInitialEffects` that checks if the store already has
data (from Jotai localStorage hydration). If empty, it fetches minimal
metadata from the new endpoint. The full metadata load continues in
parallel, eventually enriching the store with complete data.
- **SSE effects alignment**: All metadata entity types now have
corresponding SSE effects that directly patch the metadata store via
`patchMetadataStoreFromSSEEvent`.
- **Existing selectors and joining logic**:
`objectMetadataItemsWithFieldsSelector`, `viewsWithRelationsSelector`,
`pageLayoutsWithRelationsSelector` reconstruct nested data from flat
entities for components that need it.
### Loading flow
```
App mount
→ Jotai atomWithStorage hydrates store from localStorage (sync, getOnInit)
→ MinimalMetadataLoadEffect
→ Store has data? → skip (app renders immediately)
→ Store empty? → fetch minimalMetadata endpoint → populate objects + views
→ MetadataProviderInitialEffects (full metadata load, runs in parallel)
→ LazyMetadataLoadEffect (page layouts, logic functions, nav menu, etc.)
→ IsAppMetadataReadyEffect (sets isAppMetadataReady)
```
## Test plan
- [ ] Verify app loads with empty localStorage (should fetch minimal
metadata, then full)
- [ ] Verify app loads with populated localStorage (should skip minimal
fetch, render immediately)
- [ ] Verify SSE events correctly update metadata store for all entity
types
- [ ] Verify logout clears metadata store (atom reset propagates to
localStorage)
- [ ] Verify all metadata selectors return correct joined data
- [ ] CI: lint, typecheck, tests pass
## Summary
- **Removes the entire `modules/favorites/` directory** (~66 files,
~5000 lines deleted) — components, hooks, states, types, utils, tests,
and the favorite-folder-picker sub-module
- **Eliminates the dual-write pattern** where creating a favorite also
created a NavigationMenuItem — all consumers now use
`useCreateNavigationMenuItem` directly
- **Removes `IS_NAVIGATION_MENU_ITEM_EDITING_ENABLED` feature flag
checks** from ~12 files, always taking the NavigationMenuItem code path
- **Cleans up backend dual-writes** in `object-metadata.service.ts` and
`twenty-standard-application.service.ts` that were creating Favorite
records alongside NavigationMenuItems
- **Updates prefetch system** to only load NavigationMenuItems (removes
favorites prefetch effects and states)
- **Cleans up test infrastructure** — updates Storybook decorators, mock
data, and graphql mocks to remove favorites references
### What was intentionally kept
- **Backend entity definitions** (`FavoriteWorkspaceEntity`,
`FavoriteFolderWorkspaceEntity`) — these define the database schema and
need a proper database migration to remove
- **Cascade deletion listeners** — still needed to clean up existing
Favorite data in workspaces that haven't been fully migrated
- **v1.18 migration commands** — needed for workspaces upgrading from
older versions
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
## Summary
This PR upgrades Apollo Client from v3.10.0 to v4 and refactors error
handling patterns across the codebase to use a new centralized
`useSnackBarOnQueryError` hook.
## Key Changes
- **Dependency Update**: Upgraded `@apollo/client` from `^3.10.0` to
`^3.11.0` in root package.json
- **New Hook**: Added `useSnackBarOnQueryError` hook for centralized
Apollo query error handling with snack bar notifications
- **Error Handling Refactor**: Updated 100+ files to use the new error
handling pattern:
- Removed direct `ApolloError` imports where no longer needed
- Replaced manual error handling logic with `useSnackBarOnQueryError`
hook
- Simplified error handling in hooks and components across multiple
modules
- **GraphQL Codegen**: Updated codegen configuration files to work with
Apollo Client v3.11.0
- **Type Definitions**: Added TypeScript declaration file for
`apollo-upload-client` module
- **Test Updates**: Updated test files to reflect new error handling
patterns
## Notable Implementation Details
- The new `useSnackBarOnQueryError` hook provides a consistent way to
handle Apollo query errors with automatic snack bar notifications
- Changes span across multiple feature areas: auth, object records,
settings, workflows, billing, and more
- All changes maintain backward compatibility while improving code
maintainability and reducing duplication
- Jest configuration updated to work with the new Apollo Client version
https://claude.ai/code/session_019WGZ6Rd7sEHuBg9sTrXRqJ
---------
Co-authored-by: Claude <noreply@anthropic.com>
## Summary
The global `jotaiStore` singleton was shared across all tests without
reset, leaking state between test suites and causing unbounded heap
growth within each Jest worker.
- `getJestMetadataAndApolloMocksWrapper` now creates a **fresh Jotai
store per wrapper** (equivalent of RecoilRoot isolation), so each test
gets clean state
- Added `workerIdleMemoryLimit: '512MB'` to recycle workers that still
accumulate memory
- Set `maxWorkers: 3` for CI
## Performance
Measured on 48 test suites (command-menu + views + object-record hooks),
single worker:
| Metric | Before | After |
|---|---|---|
| Peak heap | **1519 MB** | **~530 MB** (-65%) |
| Final heap | **1519 MB** | **437 MB** (-71%) |
| Growth pattern | Monotonic (never freed) | Bounded sawtooth (recycled
at 512MB) |
## Summary
- update AI chat message typography and list line-height for readability
- apply richer markdown-section styling for headings, spacing,
separators, and inline code
- keep links non-underlined by default with underline on hover, using
accent11 for link color
- preserve previous AI chat table design while keeping other markdown
improvements
## Validation
- yarn eslint
packages/twenty-front/src/modules/ai/components/LazyMarkdownRenderer.tsx
packages/twenty-front/src/modules/ai/components/AIChatMessage.tsx
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
Co-authored-by: Copilot Autofix powered by AI <223894421+github-code-quality[bot]@users.noreply.github.com>
Co-authored-by: Devessier <baptiste@devessier.fr>
## Summary
This PR fixes the `tsconfig` setup in `twenty-front` so that `tsgo -p
tsconfig.json` properly type-checks all files.
### Root Cause
The previous setup used TypeScript project references with `files: []`
in the main `tsconfig.json`. When running `tsgo -p tsconfig.json`, this
checks nothing because `tsgo` requires the `-b` (build) flag for project
references, but the configs weren't set up for composite mode.
### Changes
**Simplified tsconfig architecture (4 files → 2):**
- `tsconfig.json` - All files (dev, tests, stories) for
typecheck/IDE/lint
- `tsconfig.build.json` - Production files only (excludes tests/stories)
**Removed redundant configs:**
- `tsconfig.dev.json`
- `tsconfig.spec.json`
- `tsconfig.storybook.json`
**Updated references:**
- `jest.config.mjs` → uses `tsconfig.json`
- `eslint.config.mjs` → uses `tsconfig.json`
- `vite.config.ts` → uses `tsconfig.json` for dev
**Type fixes (pre-existing errors revealed by proper typechecking):**
- Made `applicationId` optional in `FieldMetadataItem` and
`ObjectMetadataItem`
- Added missing `navigationMenuItem` translation
- Added `objectLabelSingular` to Search GraphQL query
- Fixed `sortMorphItems.test.ts` mock data
## Test plan
- [ ] Run `npx nx typecheck twenty-front` - should pass
- [ ] Run `npx nx lint twenty-front` - should work
- [ ] Run `npx nx test twenty-front` - should work
- [ ] Run `npx nx build twenty-front` - should work
- [ ] Verify IDE type checking works correctly
## Overview
This PR implements **generic many-to-many relation support** through
junction tables (also known as associative entities or join tables).
This replaces the need for hardcoded taskTarget/noteTarget logic and
provides a flexible foundation for modeling complex entity
relationships.
## Architecture
### Data Model
Many-to-many relationships are implemented using a **junction object
pattern**:
```
┌─────────┐ ┌──────────────────┐ ┌─────────┐
│ Pet │──────>│ PetRocket │<──────│ Rocket │
│ │ 1:N │ (junction) │ N:1 │ │
│ rockets ├───────┤ pet : Pet ├───────┤ │
└─────────┘ │ rocket : Rocket │ └─────────┘
└──────────────────┘
```
The junction object (PetRocket) has:
- A `MANY_TO_ONE` relation to **Pet** (the source)
- A `MANY_TO_ONE` relation to **Rocket** (the target)
The source object (Pet) has a `ONE_TO_MANY` relation pointing to the
junction, with **field settings** that specify which target field to
follow.
### Field Settings Schema
Junction configuration is stored in `FieldMetadataRelationSettings`:
```typescript
{
relationType: "ONE_TO_MANY",
// Points to the target field on the junction object
junctionTargetFieldId?: string; // For regular relations
junctionTargetMorphId?: string; // For polymorphic relations
}
```
**Two configuration modes:**
1. **`junctionTargetFieldId`** - References a specific `RELATION` field
on the junction
2. **`junctionTargetMorphId`** - References a `morphId` group for
polymorphic targets (e.g., link to Person OR Company)
### GraphQL Query Generation
When a junction relation is detected, the GraphQL fields are generated
to fetch the nested target:
```graphql
query GetPetWithRockets {
pet(id: "...") {
rockets { # ONE_TO_MANY to junction
id
rocket { # Target field on junction
id
name
__typename
}
}
}
}
```
For polymorphic junction targets:
```graphql
caretakerPerson { id, name }
caretakerCompany { id, name }
```
## Frontend Architecture
### Display Flow
1. **Detection**: `hasJunctionConfig()` checks if field has junction
settings
2. **Config Resolution**: `getJunctionConfig()` resolves junction object
metadata and target fields
3. **Record Extraction**: `extractTargetRecordsFromJunction()` extracts
target records from junction records
4. **Rendering**: Target records displayed as chips (not junction
records)
### Edit Flow
1. **Picker Opening**: Initializes the multi-record picker with:
- Searchable object types (derived from junction target fields)
- Pre-selected items (extracted from existing junction records)
2. **Selection Handling**: Manages create/delete of junction records:
- **Select**: Creates new junction record with source + target IDs
- **Deselect**: Finds and deletes the junction record
- **Optimistic Updates**: Manually updates Recoil store before API call
### Key Trade-offs
| Decision | Trade-off |
|----------|-----------|
| Junction records managed manually | More control over optimistic
updates, but requires manual cache management |
| Settings stored per-field | Flexible (same junction can power
different views), but requires UI to configure |
| Polymorphic via morphId groups | Supports N target types, but adds
query complexity |
| Feature flag gated | Safe rollout, but requires flag management |
## Backend Changes
- **Validation**: Junction target field must exist and be a valid
`MANY_TO_ONE` relation
- **Settings**: Extended `FieldMetadataRelationSettings` type with
junction fields
- **Dev Seeder**: Added sample junction objects (PetRocket,
EmploymentHistory, PetCareAgreement) for testing
## How to Test
1. Enable the `IS_JUNCTION_RELATIONS_ENABLED` feature flag
2. Create objects with junction pattern (Pet → PetRocket → Rocket)
3. Configure the junction target in field settings (advanced mode)
4. Verify:
- Display shows target objects (Rockets), not junction records
(PetRockets)
- Picker allows selecting/deselecting targets
- Changes persist correctly
https://github.com/user-attachments/assets/d04f057a-228c-4de8-af48-76bb2d72cac1
---------
Co-authored-by: Charles Bochet <charles@twenty.com>
- Create resolvers for each type of charts which needs data
transformation after the group by operation: Bar Chart, Line Chart and
Pie Chart
- Move all the utils to the backend and refactored some into services
This allows all the computation to be done in the backend, improving
performances in the frontend.
## Summary
Moves the custom ESLint rules from `tools/eslint-rules` to
`packages/twenty-eslint-rules` for better organization within the
monorepo packages structure.
## Changes
- Move `eslint-rules` from `tools/` to `packages/twenty-eslint-rules`
- Use `loadWorkspaceRules` from `@nx/eslint-plugin` to load custom rules
- Update all ESLint configs to use the `twenty/` rule prefix instead of
`@nx/workspace-`
- Update `project.json`, `jest.config.mjs` with new paths
- Update `package.json` workspaces and `nx.json` cache inputs
- Update Dockerfile reference
## Technical Details
The custom ESLint rules are now loaded using Nx's `loadWorkspaceRules`
utility which:
- Handles TypeScript transpilation automatically
- Allows loading workspace rules from any directory
- Provides a cleaner approach than the previous `@nx/workspace-`
convention
## Testing
- Verified all 17 custom ESLint rules load correctly from the new
location
- Verified linting works on dependent packages (twenty-front,
twenty-server, etc.)
# Introduction
When creating a view with v2 flag activated in production result in race
condition due to request being slow and //.
That's why we're introducing a batch create on view field here
closing https://github.com/twentyhq/core-team-issues/issues/1836
## In v2
- batch create view field endpoint is available
- frontend will target the new endpoint
## In v1
- batch create view field endpoint is not available
- frontend will stick to old fake batch view field creation loop
## Polish
- use persist view field is quite verbose as contains two data model we
could aim to create an update many view fields in order to standardize
new pattern
This big PR implements table virtualization with an offset paging,
allowing a way more fluid UX.
It is a v1 that should be improved in the future with partial data
loading and optimization of the browser display performance of a row.
But with this PR we have the solid enough technical foundation, both
frontend and backend, to get to a smooth table UX.
Fixes and improvements after first successful round of development
(needed to have main clean) :
- [x] Delete should refresh virtualized portion only and reset all table
- [x] Fix add new : top and bottom
- [x] Table empty shouldn’t show when first loading
- [x] Fix d&d
- [x] Fix sorts
- [x] Fix drag when scrolling after a full virtual page (it throws an
error)
- [x] Si update mais qu’on a un sort ou filter, alors il faut trigger le
refresh
- [x] Reset scroll position between tables
- [x] Reset scroll shadows between tables
- [x] Setup d&n for virtual list :
https://github.com/hello-pangea/dnd/blob/main/docs/patterns/virtual-lists.md
- [x] Full table re-render when entering edit mode
- [x] Clean code and prepare for merge
Fixes https://github.com/twentyhq/core-team-issues/issues/1613 that
contains other bugs to be fixed before merge
---------
Co-authored-by: Charles Bochet <charles@twenty.com>
Parallel code path to read and write core views when
IS_CORE_VIEW_ENABLED.
Migrated view key to an enum.
---------
Co-authored-by: Charles Bochet <charles@twenty.com>
Co-authored-by: Lucas Bordeau <bordeau.lucas@gmail.com>
Félix Malfait
