## Summary
- **Admin-level filtering**: New AI tab in admin panel with server-wide
model availability controls (whitelist/blacklist via
`AI_AUTO_ENABLE_NEW_MODELS`, `AI_DISABLED_MODEL_IDS`,
`AI_ENABLED_MODEL_IDS` config variables). Dedicated
`setAdminAiModelEnabled` mutation replaces frontend config-variable
manipulation. Filter dropdown to show/hide unconfigured and deprecated
models.
- **Workspace-level filtering**: Per-workspace controls with "Use best
models only" mode (curated list backed by `isRecommended` flag), or
custom whitelist/blacklist. Separate Smart/Fast model selectors with
"Best (...)" virtual options.
- **Security enforcement**: Both layers enforced at every backend
execution point — workspace update, agent create/update, chat execution.
Model ID validated against known models before config mutation. All
admin endpoints protected by `AdminPanelGuard`.
## Changes
### Backend (`twenty-server`)
- New config variables for admin-level model filtering
- `AiModelRegistryService`: `getAllModelsWithStatus()`,
`setModelAdminEnabled()` with model ID validation,
`isModelAdminAllowed()`
- `AdminPanelResolver`: `getAdminAiModels` query,
`setAdminAiModelEnabled` mutation
- `WorkspaceEntity`: new fields (`autoEnableNewAiModels`,
`disabledAiModelIds`, `enabledAiModelIds`, `useRecommendedModels`)
- `WorkspaceService`: model validation on `smartModel`/`fastModel`
updates
- `AgentResolver`: model availability checks on create/update
- `isModelAllowedByWorkspace` centralized utility
- `isRecommended` flag on model definitions
- Two TypeORM migrations
### Frontend (`twenty-front`)
- New `SettingsAdminAI` component with search, filter dropdown
(unconfigured/deprecated), and model toggle cards
- AI tab added to admin panel navigation
- `useWorkspaceAiModelAvailability` hook for workspace-level filtering
- `SettingsAIModelsTab` redesigned: merged sections, "Use best models
only" toggle, conditional available models list
- `getModelIcon`/`getModelProviderLabel` shared utilities with GraphQL
enum casing normalization
- Updated generated GraphQL types and mock data
## Test plan
- [ ] Toggle models on/off in admin panel AI tab and verify they
appear/disappear in workspace settings
- [ ] Enable "Use best models only" in workspace settings and verify
only recommended models are selectable
- [ ] Disable recommended mode and verify whitelist/blacklist toggles
work correctly
- [ ] Verify deprecated models hidden by default, shown greyed out when
filter enabled
- [ ] Verify unconfigured models hidden by default, shown disabled when
filter enabled
- [ ] Try setting a disabled model as Smart/Fast model — should be
rejected
- [ ] Try creating an agent with a disabled model — should be rejected
- [ ] Verify admin panel AI tab requires admin access
Made with [Cursor](https://cursor.com)
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
## Summary
- **Model pricing overhaul**: All model constants updated with accurate
pricing in dollars per 1M tokens, including cached input rates, cache
creation rates, and tiered >200k context pricing
- **New providers**: Added Google (Gemini 3.x), Mistral, and AWS Bedrock
as inference providers. Bedrock serves Claude Opus 4.6 and Sonnet 4.6
via AWS, with proper credential handling following the existing S3/SES
pattern
- **InferenceProvider/ModelFamily split**: Refactored `ModelProvider`
into two orthogonal enums — `InferenceProvider` (who serves the model:
auth, SDK, metadata format) and `ModelFamily` (who created it: token
counting semantics). This eliminates growing `||` chains for token
normalization checks like `excludesCachedTokens`
- **Billing improvements**: Reasoning tokens charged at output rate,
cache token discounts applied accurately, real errors thrown to Sentry
on billing failures
## Test plan
- [x] All existing unit tests updated and passing (23 tests across 3
test files)
- [x] Lint passes for both twenty-server and twenty-front
- [ ] CI checks pass
Made with [Cursor](https://cursor.com)
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
⚠️ **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>
## Context
The previous WorkspaceAuthContext was a single interface with many
optional fields, making it unclear which fields are available in
different authentication scenarios. This made the code harder to reason
about and required runtime checks scattered throughout the codebase.
## Changes
- Introduced a discriminated union type for WorkspaceAuthContext with
four specific variants:
-> UserWorkspaceAuthContext - for authenticated users
-> ApiKeyWorkspaceAuthContext - for API key authentication
-> ApplicationWorkspaceAuthContext - for application-based auth
-> SystemWorkspaceAuthContext - for system/internal operations
- Added type guard functions (isUserAuthContext, isApiKeyAuthContext,
etc.) for safe type narrowing
- Added builder utilities (buildUserAuthContext, buildApiKeyAuthContext,
etc.) to construct each context variant with proper type safety
- Refactored WorkspaceAuthContextMiddleware to use the new builders
instead of constructing a loosely-typed object
- Moved the type definition from twenty-orm/interfaces/ to
core-modules/auth/types/ for better organization
- Updated all consumers across query runners, tool providers, and
modules to use the new type location
## Notes
- I had to query User and WorkspaceMember in some parts of tool module
that were expecting userWorkspaceId but not the rest of
UserWorkspaceAuthContext (that should be required with the new proper
type otherwise it would break a lot of logic and mostly permissions with
the newly added RLS -> This is what we expect from
UserWorkspaceAuthContext and how it's done in the "normal" path in HTTP
middleware)
- WorkspaceMember is in the cache already but ideally we should move
User (And Workspace?) in the cache as well to avoid querying the DB
after each request (this is also valid for HTTP middleware when we
hydrate the Request object btw)
## Context
Introduces a middleware that automatically sets the workspace auth
context in AsyncLocalStorage for HTTP requests, making it available
throughout the request lifecycle without explicit parameter passing.
The motivation behind this change is to reduce boilerplate and simplify
the developer experience when working with workspace data in HTTP
request handlers.
The Problem (Before)
Every HTTP request handler that needed to access workspace data had to:
- Extract auth-related info from decorators (@AuthWorkspace(),
@AuthUserWorkspaceId(), etc.) in controller/resolver and pass down to
services
- Build or pass the authContext explicitly (sometimes with type
assertion which was flaky)
Then call executeInWorkspaceContext(authContext, async () => { ... })
## Changes
- Add WorkspaceAuthContextMiddleware that extracts auth context from the
request and stores it in AsyncLocalStorage
- Register middleware for GraphQL, metadata, and REST routes (runs after
hydration middlewares)
- Simplify executeInWorkspaceContext signature: fn is now the first
parameter, authContext is optional second
- If authContext is not provided, it's automatically retrieved from the
storage (set by middleware)
- Update all callers (~120 files) to use the new parameter order
- Fixes a bug in search where system auth context was used, bypassing
RLS feature.
## 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.)
## 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 migrates workflow CRUD operations to properly use the Common API
layer's authentication context, addressing the issues from the reverted
PR #15875.
The original PR was reverted because the Common API required passing
either a User or an API Key for authentication, which was problematic
for workflows. Since then, the "Application" concept was introduced in
the Common API layer, allowing for token injection in serverless
functions.
This PR leverages the "Twenty Standard Application" concept for
non-manual workflow triggers, providing a clean authentication path
without the issues of user impersonation.
## Changes
### Core Infrastructure
- **RecordCrudExecutionContext**: Replace `workspaceId` with full
`authContext`
- **WorkflowExecutionContext**: Add `authContext` field to carry
authentication info
- **ToolGeneratorContext/ToolSpecification**: Add optional `authContext`
support for tool generation
### Authentication Flow
- **WorkflowExecutionContextService**: Build appropriate auth context
based on trigger type:
- **Manual triggers**: Use user's workspace auth context with their role
permissions
- **Non-manual triggers**: Use Twenty Standard Application auth context
(bypasses permission checks or uses default serverless function role)
- **ApplicationService**: Add `findTwentyStandardApplicationOrThrow`
method to retrieve the system application
- **UserWorkspaceService**: Make relations configurable in
`getUserWorkspaceForUserOrThrow` to load only what's needed
### CRUD Services Migration
All 5 record CRUD services now receive `authContext` instead of
`workspaceId`:
- `CreateRecordService`
- `UpdateRecordService`
- `DeleteRecordService`
- `FindRecordsService`
- `UpsertRecordService`
### Workflow Actions
All record CRUD workflow actions pass `executionContext.authContext` to
the services:
- `CreateRecordWorkflowAction`
- `UpdateRecordWorkflowAction`
- `DeleteRecordWorkflowAction`
- `FindRecordsWorkflowAction`
- `UpsertRecordWorkflowAction`
### AI Agent Integration
- AI agent workflow action passes auth context to agent executor
- Tool provider and MCP protocol service support auth context
propagation
## Benefits
- ✅ Proper authentication for workflow CRUD operations via Common API
- ✅ Non-manual triggers use system application context (no user
impersonation issues)
- ✅ Manual triggers preserve user permissions correctly
- ✅ Foundation for better permission handling in automated workflows
- ✅ Cleaner separation between user-initiated and system-initiated
operations
## Related
- Reverted PR: #15875
## 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 -->
## Context
Following https://github.com/twentyhq/twenty/pull/16399
Now using the new global orm manager everywhere and returning a
GlobalDatasource/WorkspaceDatasource based on a feature flag.
This means we now need to wrap all our ORM calls within
executeInWorkspaceContext callback (at least for now) so the global
datasource can dynamically hydrate its context via the new store (the
global datasource does not store anything related to workspaces as it is
now a unique singleton). If feature flag is off it still uses local data
stored in the workspace datasource.
## 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
## Context
Deprecating TwentyORMManager in favor of TwentyORMGlobalManager
(temporarily, as this will simplify the ultimate goal to later replace
all usages with the new TwentyORMGlobalManagerV2 which will have a
similar signature)
This means this PR had to refactor a bit of code to pass down the
workspaceId when not available directly as it is now a requirement,
meaning we also deprecated scopedWorkspaceContextFactory to have a less
obscure way to fetch the workspaceId and have something more
declarative.
Step 3 will be to update TwentyORMGlobalManager to use a featureFlag
toggling and use the new GlobalWorkspaceOrmManager internally using the
new cache service
Step 4 will be to remove the feature flag and pg_pool patch
## Summary
Consolidates the AI tool provider architecture by creating a single
`ToolProviderService` as the entry point for all tool generation. This
removes multiple intermediate services and simplifies the codebase.
## Changes
### New Architecture
- **`ToolProviderService`**: Single service for all tool generation
with:
- `getTools(spec)` - Get tools by category with permissions
- `getToolByType(type)` - Get specific tool for workflow execution
- **`ToolCategory` enum**: Declarative specification of tool types:
- `DATABASE_CRUD` - Record CRUD operations
- `ACTION` - HTTP requests, email sending, article search
- `WORKFLOW` - Workflow management tools
- `METADATA` - Object/field metadata tools
- `NATIVE_MODEL` - Model-specific tools (e.g., web search)
- **`ToolSpecification` type**: Clean API for requesting tools with
permissions
### Removed
- `AiToolsModule` - No longer needed
- `ToolService` - Logic inlined into ToolProviderService
- `ToolAdapterService` - Logic inlined into ToolProviderService
- `ToolRegistryService` - Logic inlined into ToolProviderService
### Updated
- All consumers (agents, chat, MCP, workflows) now use
`ToolProviderService`
- Test files updated accordingly
## Stats
- **547 insertions, 1146 deletions** (net ~600 lines removed)
- 4 services deleted
- 1 module deleted
## Testing
- [x] Typecheck passes
- [x] Lint passes
## Summary
This PR adds a new **Metadata Builder** AI agent that specializes in
managing the workspace data model (creating objects, adding fields,
etc.).
## Changes
### New Files
- `data-model-manager-role.ts` - New standard role with `DATA_MODEL`
permission flag
- `metadata-builder-agent.ts` - New standard agent for data model
management
### Modified Files
- **ChatToolsProviderService**: Refactored to consolidate all
permission-based tools into a single `getChatTools()` method. Now
injects both workflow tools and metadata tools based on permissions.
- **AgentChatRoutingService**: Updated to use the new consolidated
`getChatTools()` method
- **AiChatModule**: Added imports for `ObjectMetadataModule` and
`FieldMetadataModule`
- **Router system prompt**: Added metadata-builder agent selection rules
with clear distinction between schema operations vs data operations
- **Metadata tools factories**: Improved error messages to show detailed
validation errors instead of generic messages
### Refactoring
- Renamed `index.ts` files to `standard-agent-definitions.ts` and
`standard-role-definitions.ts` to follow naming conventions
- Renamed exports from `standardAgentDefinitions` to
`STANDARD_AGENT_DEFINITIONS` (SCREAMING_SNAKE_CASE)
## Key Features
1. **Metadata Builder Agent** can:
- Create new custom objects
- Add fields to existing objects
- Update object and field properties
- Create relations between objects
2. **Permission-based tool injection**: Tools are automatically injected
based on the `DATA_MODEL` permission flag
3. **Improved routing**: The router now correctly distinguishes between:
- "Create an object called Project" → metadata-builder (schema)
- "Create a company called Acme" → data-manipulator (data)
4. **Better error messages**: Validation errors now show detailed
messages like:
```
Validation errors:
[objectMetadata] Name must be in camelCase format
[objectMetadata] Label is required
```
## 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 -->
Abdullah.