⚠️ **AI-generated PR — not ready for review** ⚠️
cc @FelixMalfait
---
## Changes
### System prompt improvements
- Explicit skill-before-tools workflow to prevent the model from calling
tools without loading the matching skill first
- Data efficiency guidance (default small limits, use filters)
- Pluralized `load_skill` → `load_skills` for consistency with
`load_tools`
### Token usage reduction
- Output serialization layer: strips null/undefined/empty values from
tool results
- Lowered default `find_*` limit from 100 → 10, max from 1000 → 100
### System object tool generation
- System objects (calendar events, messages, etc.) now generate AI tools
- Only workflow-related and favorite-related objects are excluded
### Context window display fix
- **Bug**: UI compared cumulative tokens (sum of all turns) against
single-request context window → showed 100% after a few turns
- **Fix**: Track `conversationSize` (last step's `inputTokens`) which
represents the actual conversation history size sent to the model
- New `conversationSize` column on thread entity with migration
### Workspace AI instructions
- Support for custom workspace-level AI instructions
---------
Co-authored-by: claude[bot] <41898282+claude[bot]@users.noreply.github.com>
https://github.com/user-attachments/assets/c4e63edb-6e98-44bc-841f-ee110ae712d4
How it works
- `manifest.json` files are now committed when apps are published to our
repo
- to display available apps, from the server, we read into our github
repo, using a pod-scoped cache
- feature flagged
- app installation will be behind permission gate MARKETPLACE_APPS
Limitations and what is yet to develop
- content and settings tabs
- installed apps tab
- app installation
- test and potentially fix reading from .manifest.json once [Reshape
manifest
structure](https://github.com/twentyhq/core-team-issues/issues/2183) is
done. additional work is expected on assets notably. (couldnt properly
do it here as manifest.json will only be committed after this pr)
- we only read in community/ folder for now - we may want tochange that
- the cache is rather artisanal for now and scoped by pod - we may want
to change that
## Summary
Fixes an issue where the AI chat would show loading shimmers
indefinitely when opened on a workspace with no conversation history.
**Root cause:** The `isLoading` state in `useAgentChat` included
`!currentAIChatThread`. On workspaces with no chat threads,
`currentAIChatThread` remained `null`, causing `isLoading` to be
permanently `true`.
**Changes:**
- Remove `!currentAIChatThread` from `isLoading` calculation in
`useAgentChat` - this state should only reflect streaming/file selection
status
- Auto-create a chat thread in `useAgentChatData` when the threads query
returns empty, ensuring a valid thread exists for the `useChat` hook to
initialize properly
- Add primary font color to empty state title for better visibility
## Test plan
1. Create a new workspace or use a workspace with no AI chat history
2. Open the AI chat
3. Verify the empty state shows (not infinite loading shimmer)
4. Send a message and verify it works correctly
Made with [Cursor](https://cursor.com)
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
## 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.)
Upgraded from 8.6.15 to 9.1.17 in two steps:
- 8.6.15 -> 9.0.0
- 9.0.0 -> 9.1.17
I had to disable `storybook-addon-cookie` since it is not supported for
Storybook 9. However, I do intend to upgrade to Storybook 10 when this
is merged, so we can replace the aforementioned add-on with this fork
specifically created to support Storybook 10 and above:
https://www.npmjs.com/package/@storybook-community/storybook-addon-cookie.
Additionally, once we upgrade to Version 10 successfully, I will start
looking into integrating the official Vitest add-on.
## Summary
This PR enables serverless functions to be exposed as AI tools, allowing
them to be used by AI agents.
### Changes
- Added new `SERVERLESS_FUNCTION` tool category
- Added `toolDescription`, `toolInputSchema`, and `toolOutputSchema`
fields to serverless functions
- Created database migration for the new schema columns
- Added tool index query and resolver for fetching available tools
- Added Settings AI page tabs (Skills, Tools, Settings) with new tools
table
- Added utility to convert tool schema to JSON schema format
- Updated frontend to display tools in the settings page
### Implementation Details
- Serverless functions can now define tool metadata (description,
input/output schemas)
- These functions are automatically registered in the tool registry
- The tool index endpoint allows querying available tools with their
schemas
- Settings page now has a dedicated Tools tab showing all available
tools
## Summary
This PR introduces a Skills system for AI agents, inspired by the [Agent
Skills specification](https://agentskills.io/specification).
## Changes
### Backend
- **SkillEntity**: New database entity with migration for storing skills
- **V2 Sync Mechanism**: Implemented FlatSkill, builders, validators,
and action handlers following the v2 flat entity pattern
- **Standard Skills**: Pre-defined skills (workflow-building,
data-manipulation, dashboard-building, metadata-building, research,
code-interpreter, xlsx, pdf, docx, pptx)
- **GraphQL API**: CRUD operations for skills with proper guards and
permissions
- **Workspace Cache**: Integrated skills into the workspace cache system
### Frontend
- **Skills Table**: Searchable table in AI settings showing all skills
- **Skill Form**: Create/edit page with Label (primary), Description,
and Content (markdown editor)
- **API Name**: Following existing patterns, name is derived from label
with advanced settings toggle for custom API names
- **Standard vs Custom**: Standard skills are read-only, custom skills
can be edited/deleted
## Key Design Decisions
- Skills are stored in the database (Salesforce-like approach) rather
than files
- Name is derived from Label by default (isLabelSyncedWithName pattern)
- Skills reference functions/files via @ mentions in markdown content
rather than explicit relations
- Standard skills are synced from code, custom skills are created via UI
## Screenshots
Skills table and form UI follow existing settings patterns.
## Testing
- [x] Lint passes
- [x] Typecheck passes
- [ ] CI tests
## Summary
This PR ensures usage alerts are created for all billing scenarios.
## Background
Per [Stripe
documentation](https://docs.stripe.com/billing/subscriptions/usage-based/alerts),
usage alerts are **one-time per customer** - they trigger once and only
consider usage reported after the alert is created. This means we need
to create a new alert whenever:
1. ✅ Subscription is created (trial) - Already implemented
2. ✅ Trial ends (user becomes Active subscriber) - **Added in this PR**
3. ✅ Credit tier changes (upgrade) - **Added in this PR**
4. ✅ **New billing cycle starts** - **Critical fix in this PR!**
## The Issue
Previously, the invoice webhook reset `hasReachedCurrentPeriodCap =
false` at cycle end, but didn't create a new alert. This meant after the
first billing period, there was no alert to trigger and users could
exceed their limit without being blocked.
## Changes
### 1. Invoice Webhook (`billing-webhook-invoice.service.ts`)
When invoice is finalized for `subscription_cycle`:
- Reset `hasReachedCurrentPeriodCap = false` ✅ (already done)
- **NEW**: Create alert at the current tier cap
### 2. End Trial Period (`billing-subscription.service.ts`)
In `endTrialPeriod()`:
- **NEW**: Create alert at the paid tier cap (not trial cap)
### 3. Credit Tier Upgrades (`billing-subscription.service.ts`)
In `changeMeteredPrice()`, when upgrading immediately (not scheduled for
period end):
- **NEW**: Reset `hasReachedCurrentPeriodCap = false`
- **NEW**: Create alert at new tier cap
### 4. Alert Title Update (`stripe-billing-alert.service.ts`)
Changed from "Trial usage cap" to "Usage cap" since alerts are now used
for all scenarios.
## Stripe Alert Limits
- Max 25 alerts per meter+customer combination
- Alerts only evaluate usage reported after creation
- One-time alerts trigger once per customer
With monthly billing cycles + occasional tier changes, we should stay
well under the 25 alert limit.
## Testing
- TypeScript typechecking passes
- Backend services properly inject new dependencies
## Summary
This PR adds user-friendly error handling for AI chat features,
specifically for **billing credits exhausted** and **API key not
configured** errors.
## Changes
### Backend
- Added `BILLING_CREDITS_EXHAUSTED` exception code with 402 status
- Added `API_KEY_NOT_CONFIGURED` exception code with 503 status
- Added billing check before AI chat streaming in
`agent-chat.controller.ts`
- Added error code to HTTP exception response body for frontend error
type detection
- Created `AgentRestApiExceptionFilter` for agent-specific errors
### Frontend
- Created `AIChatBanner` - reusable banner component for error/warning
messages
- Created `AIChatCreditsExhaustedMessage` - shows upgrade prompts based
on user permissions
- Created `AIChatApiKeyNotConfiguredMessage` - shows configuration
guidance with docs link
- Created `AIChatErrorRenderer` - encapsulates error type switching
logic (fixes nested ternary)
- Created `AIChatStandaloneError` - displays errors when there are no
messages
- Split `aiChatErrorUtils.ts` into separate files (1 export per file):
- `AIChatErrorCode.ts`
- `extractErrorCode.ts`
- `isAIChatErrorOfType.ts`
- `isBillingCreditsExhaustedError.ts`
- `isApiKeyNotConfiguredError.ts`
- Added comprehensive test coverage (27 tests)
### Other
- Updated trial period banner messaging
## Testing
- All lint checks pass
- All 27 new tests pass
- TypeScript typecheck passes
## Summary
This PR enforces the use of `@/` alias for imports instead of relative
parent imports (`../`).
## Changes
### ESLint Configuration
- Added `no-restricted-imports` pattern in `eslint.config.react.mjs` to
block `../*` imports with the message "Relative parent imports are not
allowed. Use @/ alias instead."
- Removed the non-working `import/no-relative-parent-imports` rule
(doesn't work properly in ESLint flat config)
### VS Code Settings
- Added `javascript.preferences.importModuleSpecifier: non-relative` to
`.vscode/settings.json` (TypeScript setting was already there)
### Code Fixes
- Fixed **941 relative parent imports** across **706 files** in
`packages/twenty-front`
- All `../` imports converted to use `@/` alias
## Why
- Consistent import style across the codebase
- Easier to move files without breaking imports
- Better IDE support for auto-imports
- Clearer understanding of where imports come from
## Summary
This PR adds the `lingui/no-unlocalized-strings` ESLint rule to detect
untranslated strings and fixes translation issues across multiple
components.
## Changes
### ESLint Configuration (`eslint.config.react.mjs`)
- Added comprehensive `ignore` patterns for non-translatable strings
(CSS values, HTML attributes, technical identifiers)
- Added `ignoreNames` for props that don't need translation (className,
data-*, aria-*, etc.)
- Added `ignoreFunctions` for console methods, URL APIs, and other
non-user-facing functions
- Disabled rule for debug files, storybook, and test files
### Components Fixed (~19 files)
- Object record components (field inputs, pickers, merge dialogs)
- Settings components (accounts, admin panel)
- Serverless function components
- Record table and title cell components
## Status
🚧 **Work in Progress** - ~124 files remaining to fix
This PR is being submitted as draft to allow progressive fixing of
remaining translation issues.
## Testing
- Run `npx eslint "src/**/*.tsx"` in `packages/twenty-front` to check
remaining issues
## Summary
- Add code interpreter tool that enables AI to execute Python code for
data analysis, CSV processing, and chart generation
- Support for both local (development) and E2B (sandboxed production)
execution drivers
- Real-time streaming of stdout/stderr and generated files
- Frontend components for displaying code execution results with
expandable sections
## Code Quality Improvements
- Extract `getMimeType` to shared utility to reduce code duplication
between drivers
- Fix security issue: escape single quotes/backslashes in E2B driver env
variable injection
- Add `buildExecutionState` helper to reduce duplicated state object
construction
- Add `DEFAULT_CODE_INTERPRETER_TIMEOUT_MS` constant for consistency
- Fix lingui linting warning and TypeScript theme errors in frontend
## Test Plan
- [ ] Test code interpreter with local driver in development
- [ ] Test code interpreter with E2B driver in production environment
- [ ] Verify streaming output displays correctly in chat UI
- [ ] Verify generated files (charts, CSVs) are uploaded and
downloadable
- [ ] Test file upload flow (CSV, Excel) triggers code interpreter
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> Updates generated i18n catalogs for Polish and pseudo-English, adding
strings for code execution/output (code interpreter) and various UI
messages, with minor text adjustments.
>
> - **Localization**:
> - **Generated catalogs**: Refresh `locales/generated/pl-PL.ts` and
`locales/generated/pseudo-en.ts`.
> - Add strings for code execution/output (e.g., code, copy code/output,
running/waiting states, download files, generated files, Python code
execution).
> - Include new UI texts (errors, prompts, menus) and minor text
corrections.
> - No changes to `pt-BR`; other files unchanged functionally.
>
> <sup>Written by [Cursor
Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit
befc13d02c21e5a6647bc1aa6daa2a89f60b7ef8. This will update automatically
on new commits. Configure
[here](https://cursor.com/dashboard?tab=bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
## Summary
- Add a context usage indicator to the AI chat interface inspired by
Vercel's AI SDK Context component
- Display token consumption, context window utilization percentage, and
estimated cost in credits
- Show a circular progress ring with percentage, revealing detailed
breakdown on hover
## Changes
### Backend
- Stream usage metadata (tokens, model config) via `messageMetadata`
callback in `agent-chat-streaming.service.ts`
- Return model config from `chat-execution.service.ts`
- Add usage and model types to `ExtendedUIMessage` metadata
### Frontend
- New `ContextUsageProgressRing` component - circular SVG progress
indicator
- New `AIChatContextUsageButton` component with hover card showing:
- Progress bar with used/total tokens
- Input/output token counts with credit costs
- Total credits consumed
- Track cumulative usage in Recoil state (`agentChatUsageState`)
- Reset usage when creating new chat thread
- Integrate button into `AIChatTab`
## Test plan
- [ ] Open AI chat and send a message
- [ ] Verify the context usage button appears with percentage
- [ ] Hover over the button to see detailed breakdown
- [ ] Verify credits are calculated correctly
- [ ] Create a new chat thread and verify usage resets to 0
## Summary
- Add latest AI models from OpenAI (GPT-4.1, o3, o4-mini), Anthropic
(Claude 4.5 Opus/Sonnet/Haiku), and xAI (Grok 4.1)
- Mark deprecated models (GPT-4o, GPT-4o-mini, GPT-4-turbo, Claude Opus
4, Claude Sonnet 4) with a `deprecated` flag
- Split AI models into separate files per provider for better
maintainability
- Support comma-separated default model lists for automatic fallback
across providers (works out of the box for self-hosters regardless of
which provider they configure)
- Filter deprecated models from dropdown selection while keeping them
functional for existing agents
## Changes
### New Models Added
| Provider | Models |
|----------|--------|
| OpenAI | gpt-4.1, gpt-4.1-mini, o3, o4-mini |
| Anthropic | claude-opus-4-5, claude-sonnet-4-5, claude-haiku-4-5 |
| xAI | grok-4-1-fast-reasoning |
### Deprecated Models
- gpt-4o, gpt-4o-mini, gpt-4-turbo (OpenAI)
- claude-opus-4-20250514, claude-sonnet-4-20250514 (Anthropic)
### Config Changes
Default model configs now support comma-separated fallback lists:
-
`DEFAULT_AI_SPEED_MODEL_ID=gpt-4.1-mini,claude-haiku-4-5-20251001,grok-3-mini`
-
`DEFAULT_AI_PERFORMANCE_MODEL_ID=gpt-4.1,claude-sonnet-4-5-20250929,grok-4`
## Test plan
- [x] Unit tests pass
- [x] Typecheck passes
- [x] Lint passes
- [ ] Verify deprecated models don't appear in model dropdowns
- [ ] Verify agents with deprecated models still work correctly
- [ ] Verify default model fallback works when only one provider is
configured
## Summary
Adds a new **VIEW** tool category for the AI chat, enabling it to work
with views:
- **get-views**: List views in the workspace, optionally filtered by
object metadata ID
- **get-view-query-parameters**: Convert a view's filters and sorts into
GraphQL query parameters that can be passed to existing `find_*` data
tools
- **create-view**, **update-view**, **delete-view**: CRUD operations for
view management
### Key design decisions
1. **No pagination duplication**: Instead of creating a
`find-records-from-view` tool that would duplicate pagination logic,
`get-view-query-parameters` returns filter/sort parameters that the AI
can pass to existing record-fetching tools.
2. **Permission model**:
- Read tools (get-views, get-view-query-parameters) are available to all
users
- Write tools require the `VIEW` permission
- UNLISTED views can only be modified by their creator
3. **Leverages existing utilities**: Uses
`computeRecordGqlOperationFilter` from `twenty-shared` for filter
conversion.
### Files changed
- Added `ViewToolProvider`, `ViewToolsFactory`, and
`ViewQueryParamsService`
- Added `VIEW` to `ToolCategory` enum and tool registry
- Updated `chat-execution.service.ts` to include view tools in the
catalog and pass viewId in browsing context
- Extracted shared `formatValidationErrors` utility to reduce
duplication
## Test plan
- [x] Unit tests for `ViewToolsFactory`
- [x] Unit tests for `ViewQueryParamsService`
- [x] Lint and typecheck pass
## Summary
- Add `BrowsingContext` type to automatically pass what the user is
currently viewing (recordPage or listView) to the AI chat
- Simplify context architecture: remove toggleable context UI, make it
automatic and invisible to the user
- Fix tool loading: add `unionOf` handling in
`getDatabaseToolsForObject` and fix regex ordering so `find_one_*` tools
are properly registered
- Use plural names for find tools (`find_people` vs `find_one_person`)
for better semantics
- Clean up unused components and states
## Changes
### Frontend
- New `BrowsingContext` type and `useGetBrowsingContext` hook to gather
context from Recoil state
- Simplified `useAgentChat` to use the new browsing context
- Removed toggleable context UI components
(`AgentChatContextRecordPreview`, `SendMessageWithRecordsContextButton`,
etc.)
- Removed `isAgentChatCurrentContextActiveState`
### Backend
- New `BrowsingContextType` for recordPage and listView contexts
- Updated `ChatExecutionService` to build context from browsing context
- Fixed `tool-registry.service.ts`:
- Added `unionOf` handling in permission config
- Fixed regex ordering (`find_one` before `find`) so tools load
correctly
- Use plural names for search tools (`find_people` instead of
`find_person`)
## Test plan
- [x] Typecheck passes
- [x] Lint passes
- [ ] Test AI chat on record page - should show context in system prompt
- [ ] Test AI chat on list view - should show view name and filters
- [ ] Test `find_one_*` tools now load correctly
- [ ] Test `find_*` tools use plural naming
## Summary
This PR significantly simplifies the AI chat architecture by removing
complex routing/planning mechanisms and introduces clickable record
links in AI responses.
## Changes
### AI Chat Architecture Simplification
- **Removed** the entire `ai-chat-router` module (~850 lines) including:
- Strategy decider service
- Plan generator service
- Complex routing logic
- **Removed** agent execution planning services (~700 lines):
- `agent-execution.service.ts`
- `agent-plan-executor.service.ts`
- `agent-tool-generator.service.ts`
- **Added** centralized `ToolRegistryService` for tool management:
- Builds searchable tool index (database, action, workflow tools)
- Provides tool lookup by name
- Supports agent search for loading expertise
- **Added** `ChatExecutionService` as simple replacement:
- Includes full tool catalog in system prompt
- Pre-loads common tools (find/create/update for company, person,
opportunity, task, note)
- Uses `load_tools` mechanism for dynamic tool activation
- Enables native web search by default
### Record References in AI Responses
- Added `recordReferences` field to tool outputs for create, find, and
update operations
- Implemented `[[record:objectName:recordId:displayName]]` syntax for AI
to reference records
- Created `RecordLink` component that renders clickable chips with
object icons
- Integrated record link parsing into the markdown renderer
- Users can now click directly on created/found records in AI responses
### Workflow Agent Fixes
- Fixed cache invalidation issue when creating agents in workflows
- Added default prompt for workflow-created agents to prevent validation
errors
- Relaxed agent validation to only check properties being updated (not
all required properties)
### Code Quality Improvements
- Extracted `getRecordDisplayName` utility that mirrors frontend's
`getLabelIdentifierFieldValue` logic
- Uses object metadata to determine the correct label identifier field
- Handles `FULL_NAME` composite type for person/workspaceMember objects
- Shared across create, find, and update record services
## Net Impact
- **~1,200 lines deleted** (complex routing/planning code)
- **~500 lines added** (simpler tool registry + record links)
- Significantly reduced code complexity
- Better tool discovery through full catalog in system prompt
- Improved UX with clickable record references
## Testing
- Typecheck passes
- Lint passes
- Manual testing of AI chat with record creation and linking
## Changes
### 1. Fixed streaming lag with throttle parameter
**Problem:** The AI chat was experiencing lag during message streaming
because the messages array was being updated too frequently, causing all
messages to re-render too quickly.
**Solution:** Added the `experimental_throttle: 100` parameter to the
`useChat` hook configuration. This throttles message updates during
streaming to prevent excessive re-renders and improve performance.
### 2. Cleaned up useAgentChat hook return values
**Context:** The `useAgentChat` hook primarily returns values from the
underlying `useChat` hook, so there wasn't significant room for
improvement regarding the "umbrella hook" pattern. However, some
unnecessary values were being returned that weren't needed.
**Solution:**
- Removed `input` and `handleInputChange` from the `useAgentChat` hook
return. These weren't needed since input state is already managed
directly via Recoil state (`agentChatInputState`) in components.
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> Throttle message streaming updates and switch AI chat input management
from context to Recoil; minor scroll behavior tweak.
>
> - **AI Chat performance**:
> - Add `experimental_throttle: 100` to `useChat` in `useAgentChat` to
reduce re-render frequency during streaming.
> - **State management**:
> - Migrate input handling to Recoil via `agentChatInputState`; remove
`input` and `handleInputChange` from `AgentChatContext` and
`useAgentChat` returns.
> - Update `AIChatTab` and `SendMessageButton` to read/write input from
Recoil and adjust hotkeys/disabled state accordingly.
> - **UX behavior**:
> - Remove smooth scroll behavior in `useAgentChatScrollToBottom`.
>
> <sup>Written by [Cursor
Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit
911b341ea1f7c63a4e1e27667da119d9a2960b78. This will update automatically
on new commits. Configure
[here](https://cursor.com/dashboard?tab=bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
## Summary
This PR introduces a comprehensive agent evaluation system and refactors
the AI module structure for better organization.
## Key Changes
### 🎯 Agent Evaluation System
- Added **Agent Turn Evaluation** entities, DTOs, and database schema
- New GraphQL mutations: `evaluateAgentTurn` and `runEvaluationInput`
- Added `evaluationInputs` field to Agent entity for storing test inputs
- New `AgentTurnGraderService` for automatic turn evaluation
- Added evaluation UI with new **Evals** and **Logs** tabs in agent
detail pages
### 🏗️ Entity & Module Refactoring
- Renamed `AgentChatMessage` → `AgentMessage` for clarity
- Consolidated chat entities: `AgentMessage`, `AgentTurn`, and
`AgentChatThread`
- Reorganized AI modules under `ai/` subdirectory structure
- Updated imports across codebase to reflect new module paths
### 🤖 New Agents & Roles
- Added **Dashboard Builder Agent** for dashboard creation and
management
- Added **Dashboard Manager Role** with appropriate permissions
- Updated role permissions to be more granular (users vs agents vs API
keys)
### 🔐 Permission System Updates
- Added `HTTP_REQUEST_TOOL` permission flag
- Updated Workflow Manager role permissions (restricted tool access)
- Enhanced permission flag types to differentiate between user/agent/API
key contexts
- Added `isRelevantForAgents`, `isRelevantForApiKeys`,
`isRelevantForUsers` to permission flags
### 📨 Message Role Enhancement
- Added `system` role to `AgentMessageRole` enum (alongside
user/assistant)
- Updated message handling to support system prompts
### 🎨 UI/UX Improvements
- New tabs in agent detail: **Evals** and **Logs**
- Added turn detail page: `/ai/agents/:agentId/turns/:turnId`
- Fixed text overflow in `SettingsListItemCardContent`
- Updated role applicability labels ("Assignable to Workspace Members")
### 🛠️ Technical Improvements
- Fixed Zod schema validation for UUID and Date fields (use string
validators)
- Updated `ToolRegistryService` to properly register HTTP tool with
permission flag
- Enhanced error handling in agent execution services
- Updated database migrations for new entity schema
## Database Migrations
- `1764210000000-add-system-role-to-agent-message.ts`
- `1764220000000-add-evaluation-inputs-to-agent.ts`
- `1764200000000-add-agent-turn-evaluation.ts`
- `1764100000000-refactor-agent-chat-entities.ts`
## Testing
- [ ] Agent evaluation flow tested
- [ ] Dashboard Builder agent tested
- [ ] Permission system validated
- [ ] UI tabs and navigation tested
- [ ] Database migrations run successfully
## Breaking Changes
⚠️ **Entity Rename**: `AgentChatMessage` renamed to `AgentMessage` -
GraphQL queries need updating
## Related Issues
<!-- Link any related issues here -->
## Screenshots
<!-- Add screenshots if applicable -->
## Overview
This PR replaces the dynamic agent handoff system with a more
predictable planning-based router that decides upfront how to handle
multi-agent coordination.
## Major Changes
### 🔄 Architecture Shift: Handoffs → Planning
**Removed:**
- `AgentHandoffEntity` and handoff tracking system
- `AgentHandoffService` and `AgentHandoffExecutorService`
- Dynamic agent-to-agent transfers during execution
- Handoff tool generation and description templates
**Added:**
- `AiRouterService` with two strategies: `simple` (single agent) and
`planned` (multi-agent)
- `AgentPlanExecutorService` for executing multi-step plans
- Plan validation (cycle detection, dependency resolution)
- `UnifiedRouterResult` type with discriminated union
### 🤖 New Standard Agents
Added two new specialized agents:
- **Researcher Agent**: Web search, fact-finding, competitive
intelligence
- **Code Agent**: TypeScript function generation for serverless
workflows
### 🏗️ Router Refactoring (Latest)
Split router responsibilities into focused services:
- `AiRouterStrategyDeciderService`: Decides simple vs planned strategy
- `AiRouterPlanGeneratorService`: Generates and validates execution
plans
- `AiRouterService`: Coordinates between services (reduced from 426→275
lines)
### ⚙️ Configuration Improvements
- Added `outputStrategy` to agent definitions (`direct` vs `synthesize`)
- Removed hardcoded special cases for workflow-builder
- Added `plannerModel` field to workspace entity
- Increased `MAX_STEPS` from 10 to 25 for complex workflows
### 📝 Agent Prompt Refinements
Significantly simplified prompts for better clarity:
- Workflow Builder: 51→36 lines
- Helper: 49→28 lines
- Data Manipulator: Enhanced with sorting guidance
### 🔍 Enhanced Debugging
- Plan reasoning and step count in data message parts
- Router debug info with token usage tracking
- Better logging throughout execution pipeline
## Benefits
1. **Simpler Mental Model**: Router decides upfront vs dynamic transfers
2. **Better Predictability**: Users see the plan before execution
3. **Cleaner Architecture**: SRP with focused services
4. **Configuration Over Code**: Agent behavior via config, not hardcoded
logic
5. **Plan Validation**: Catches invalid dependencies and cycles
## Migration Notes
- Database migration removes `agentHandoff` table
- Adds `plannerModel` column to workspace table
- No API breaking changes (agent endpoints unchanged)
## Testing
- Integration tests updated to remove handoff dependencies
- Agent tool test utilities simplified
- Plan validation covered by new logic
## Next Steps (Future PRs)
- Parallel execution of independent plan steps
- Dynamic re-planning based on results
- Plan caching for common routing patterns
- Error recovery strategies in plan executor
[Figma
Design](https://www.figma.com/design/xt8O9mFeLl46C5InWwoMrN/Twenty?node-id=81380-344641&t=FpjWNOK2gZuDQQfr-0)
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> Adds a side panel layout for the Command Menu, routes modals into a
local container, updates the top bar and context chips, and standardizes
small button sizes.
>
> - **Command Menu**:
> - **Side Panel Layout**: Introduces `CommandMenuSidePanelLayout` with
animated width, hosts `CommandMenuRouter`, and provides a modal
container via `ModalContainerContext`.
> - **Top Bar**: Redesign (`CommandMenuTopBar`) with back icon, optional
AI sparkles action, compact height
(`COMMAND_MENU_SEARCH_BAR_HEIGHT=40`), and updated placeholder.
> - **Context Chips**: Adds `CommandMenuLastContextChip` and
`CommandMenuRecordInfo`; extends `CommandMenuContextChip` with `page`
prop; updates `CommandMenuContextChipGroups` to render last chip as
record info when applicable.
> - **Container Simplification**: `CommandMenuContainer` simplified to
just provide contexts and `AgentChatProvider`.
> - **Modal System**:
> - Adds `ModalContainerContext` and updates `Modal` to portal into
provided container; `Modal.Backdrop` supports `isInContainer`.
> - Updates usages (e.g., `UserOrMetadataLoader`, `ActionModal`) to
align with new modal behavior.
> - **Page Integration**:
> - Replaces `PageBody` with `CommandMenuSidePanelLayout` in
`RecordShowPage` and `RecordIndexContainerGater`.
> - Removes global `CommandMenuRouter` from `DefaultLayout` (keeps
keyboard shortcuts).
> - **UI/Styling**:
> - Standardizes several buttons to `size="small"` (e.g., command
actions, open record, options, reply, workflow footer).
> - Adjusts `ShowPageSubContainer` styling when rendered inside command
menu.
> - Storybook tests updated for new placeholder text.
>
> <sup>Written by [Cursor
Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit
81fcaa145618a2fa1c3e11e2dc88fe832c51374c. This will update automatically
on new commits. Configure
[here](https://cursor.com/dashboard?tab=bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
Co-authored-by: Lucas Bordeau <bordeau.lucas@gmail.com>
Co-authored-by: Devessier <baptiste@devessier.fr>
Co-authored-by: Aman Raj <92664006+araj00@users.noreply.github.com>
Co-authored-by: Etienne <45695613+etiennejouan@users.noreply.github.com>
Co-authored-by: Paul Rastoin <45004772+prastoin@users.noreply.github.com>
## Summary
This PR adds configurable response format support for AI agents,
allowing them to return either plain text or structured JSON data based
on a defined schema.
## Key Features
### 1. Agent Response Format Configuration
- Added `AgentResponseFormat` type supporting:
- `text`: Returns plain text responses (default)
- `json`: Returns structured JSON based on defined schema
- New `AgentResponseSchema` type moved to `twenty-shared/ai` for sharing
between frontend/backend
### 2. Settings UI
- New `SettingsAgentResponseFormat` component for configuring response
format
- Visual schema builder for defining JSON output structure
- Real-time validation and preview
- Integrated into agent settings tab
### 3. Workflow Integration
- AI Agent workflow action automatically uses agent's configured
response format
- Output schema dynamically generated from agent's response format
- Workflow variable picker shows structured fields for JSON responses
- Backward compatible with existing text-only agents
### 4. Backend Implementation
- Added `convertAgentSchemaToZod` utility to validate JSON responses
- Agent executor service handles both text and JSON generation
- Automatic agent creation/cloning when adding AI agent steps to
workflows
- Unique agent naming with conflict resolution
### 5. Database Migration
- Migration `1763622159656-update-agent-response-format.ts`
- Sets default `responseFormat` to `{"type":"text"}` for existing agents
- Updated all standard agents with proper response format
## Changes by Module
### Frontend (`twenty-front`)
- 🆕 `AgentResponseFormat` type
- 🆕 `SettingsAgentResponseFormat` component
- ✏️ Updated `WorkflowEditActionAiAgent` to support response format
configuration
- 🗑️ Removed deprecated `useAiAgentOutputSchema` hook and
`AiAgentOutputSchema` type
### Backend (`twenty-server`)
- 🆕 `AgentResponseFormat` type in agent entity
- 🆕 `convertAgentSchemaToZod` utility for schema validation
- ✏️ Updated `AiAgentExecutorService` to handle both text and JSON
generation
- ✏️ Updated `WorkflowSchemaWorkspaceService` to generate output schema
from agent config
- ✏️ Enhanced `WorkflowVersionStepOperationsWorkspaceService` with agent
creation/cloning
- 🆕 Agent naming constants for conflict resolution
### Shared (`twenty-shared`)
- 🆕 `AgentResponseSchema` type
- 🆕 `ModelConfiguration` type moved to shared package
- Updated exports in `ai/index.ts`
## Code Quality
- Removed useless comments following code style guidelines
- All linter checks passed
- Type-safe implementation with proper TypeScript types
## Testing
- ✅ Database migration tested
- ✅ Agent creation/cloning in workflows verified
- ✅ Response format switching (text ↔ JSON) validated
- ✅ Backward compatibility with existing agents confirmed
## Migration Notes
- Existing agents will have `responseFormat: {type: 'text'}` set
automatically
- No breaking changes - all existing functionality preserved
- Agents can be updated to use JSON format through settings UI
**Root Cause**
Many-to-one relation filters were being exposed under the relation name
(e.g., `company`) instead of the corresponding foreign-key attribute
(e.g., `companyId`).
**Change**
- Detect many-to-one relation metadata and remap those filter keys to
`<fieldName>Id`.
- Removed some unused code unrelated to this fix.
## Overview
This PR strengthens our permission system by introducing more granular
role-based access control across the platform.
## Changes
### New Permissions Added
- **Applications** - Control who can install and manage applications
- **Layouts** - Control who can customize page layouts and UI structure
- **AI** - Control access to AI features and agents
- **Upload File** - Separate permission for file uploads
- **Download File** - Separate permission for file downloads (frontend
visibility)
### Security Enhancements
- Implemented whitelist-based validation for workspace field updates
- Added explicit permission guards to core entity resolvers
- Enhanced ESLint rule to enforce permission checks on all mutations
- Created `CustomPermissionGuard` and `NoPermissionGuard` for better
code documentation
### Affected Components
- Core entity resolvers: webhooks, files, domains, applications,
layouts, postgres credentials
- Workspace update mutations now use whitelist validation
- Settings UI updated with new permission controls
### Developer Experience
- ESLint now catches missing permission guards during development
- Explicit guard markers make permission requirements clear in code
review
- Comprehensive test coverage for new permission logic
## Testing
- ✅ All TypeScript type checks pass
- ✅ ESLint validation passes
- ✅ New permission guards properly enforced
- ✅ Frontend UI displays new permissions correctly
## Migration Notes
Existing workspaces will need to assign the new permissions to roles as
needed. By default, all new permissions are set to `false` for non-admin
roles.
## Summary
Reverts commits afc518a, ae22e64, and 800b5b5 that refactored
`useAgentChat` to address umbrella hook pattern feedback.
## Issues Introduced by Refactoring
The refactoring broke several critical functionalities:
1. **Streaming fails on thread switch** - Messages don't stream properly
when switching between threads
2. **Messages lost on tab close** - When the Ask AI tab is closed, the
request is lost instead of continuing in the background
3. **Blank chat requiring force-reload** - Chats often appear blank and
require switching to another chat to force a reload (closes
[#1771](https://github.com/twentyhq/core-team-issues/issues/1771))
## Root Cause
After extensive debugging, it appears **multiple instances of `useChat`
don't work well together**. The refactored architecture inadvertently
created scenarios where multiple `useChat` instances interfere with each
other.
## Resolution
Reverting to restore functionality. The umbrella hook pattern
optimization needs a different architectural approach that doesn't rely
on multiple `useChat` instances.
## Follow-up
While the umbrella hook feedback is valid, we need to rethink the
implementation strategy:
- Find an alternative to multiple `useChat` instances
- Possibly consolidate chat state management differently
Adds intelligent routing system that automatically selects the best
agent for user queries based on conversation context.
### Changes:
- Added `routerModel` column to workspace table for configurable router
LLM selection
- Implemented `RouterService` with conversation history analysis and
agent matching logic
- Created router settings UI in AI Settings page with model dropdown
- Removed agent-specific thread associations - threads are now
agent-agnostic
- Added real-time routing status notification in chat UI with shimmer
effect
- Removed automatic default assistant agent creation
- Renamed GraphQL operations from agent-specific to generic (e.g.,
`agentChatThreads` → `chatThreads`)
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
Co-authored-by: Félix Malfait <felix@twenty.com>
# Migrate Attachment Author to CreatedBy Field
**Twill Task**: https://twill.ai/twentyhq/ENG/tasks/7
## Summary
This PR implements a migration to transition the `Attachment` object
from using an `author` relation field to using the standard `createdBy`
field, addressing issue
https://github.com/twentyhq/core-team-issues/issues/1594.
## Changes
- **Added migration command**
(`1-8-migrate-attachment-author-to-created-by.command.ts`):
- Migrates existing attachment data to use `createdBy` instead of
`author`
- Ensures data integrity during the transition to the standard field
pattern
- **Updated Attachment workspace entity**:
- Added `createdBy` relation field to the `Attachment` standard object
- Registered new field ID in `standard-field-ids.ts` constants
- **Integrated migration into upgrade pipeline**:
- Added migration module for version 1.8
- Registered in the main upgrade version command module
This change aligns the `Attachment` object with Twenty's standard field
conventions by using the built-in `createdBy` field instead of a custom
`author` field.
---
Fixes https://github.com/twentyhq/core-team-issues/issues/1594
---------
Co-authored-by: Twill <agent@twill.ai>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
Co-authored-by: Félix Malfait <felix@twenty.com>
Closes [#1583](https://github.com/twentyhq/core-team-issues/issues/1583)
- Removed `messageId` column from `File` table and its references in
code (unrelated to this PR)
- Updated AI chat to use React Context API with persistent provider in
`CommandMenuContainer`
- Converted all AI chat component states to regular Recoil atoms (no
instance context needed)
---------
Co-authored-by: Félix Malfait <felix@twenty.com>
This PR refactors table resize with the new layout using flex-wrap to
manage all the table with divs.
Due to flex-wrap, we have to manually compute some filling divs like the
last column filler div or the empty table container, otherwise the
header can wrap and display on two lines, which is the only real issue
with this implementation with wrap.
Remember that **using flex-wrap is necessary** to handle z-index as we
want for the scrolling and hovered portal cell interaction UX. Otherwise
we lose the capacity to have z-index fine-tuning on very specific edge
cases.
To handle the resize, the most performant solution was to create CSS
variables for each column width on the table, creating a class for each
column, then assigning each class to all divs that are in the same
column, using the index of the array or the fixed known position index
for that.
This PR also refactors a lot of small details around the resizing, the
naming of the component and the DX.
It notably introduces a new hook : `useHTMLElementByIdWhenAvailable`,
that is very relies on MutationObserver API to get a native HTML element
as soon as it is rendered.
---------
Co-authored-by: Charles Bochet <charles@twenty.com>
Previous refacto was creating output schema V2 which has more specific
schemas based on the step type. Before we were using one common schema,
which was too complex when searching for variable informations.
This PR migrate the deprecated schemas and remove the old code:
- mark previous `BaseOutputSchema` as deprecated
- remove other previous schemas
- use V2 everywhere
- icon should not be stored in schema. Instead it should be generated
based on the fieldmetadata or the item type
The goal of this PR is to test if Twenty can support a large number of
members, a question which was raised by a large company that is
considering moving away from Salesforce.
I was expecting the currentWorkspaceMembersState to cause a lot more
issue. It would be very hard to get rid of it in the context of actors,
I think that would require a big refactoring. I thought we'd have to do
it but it turns out the perf are pretty good. One thing we need to
improve is the pagination on the roles page, we'll wait for @Bonapara to
update that
Félix Malfait