Several fixes after discussing with @BOHEUS
- set applicationManifest env key optional
- fix server local serverless function logging (introduces a new env
variable `SERVERLESS_LOGS_ENABLED` defaulting to false)
Implements permission intersection (AND logic) to prevent permission
escalation when agents act on behalf of users.
### Changes:
- **Permission Intersection**: Operations requiring both user AND agent
permissions
- **RoleContext Type**: Unified type supporting single `roleId` or
multiple `roleIds` for intersection
- **CRUD Services**: Updated to accept `roleContext` for granular
permission control
- **Agent Integration**: Chat agents now use user + agent role
intersection for all operations
- **ORM Layer**: Enhanced `getRepository` to support multi-role
permission checks
### Related:
- Part 2 of ["Acting on behalf of user" concept
PR](https://github.com/twentyhq/twenty/pull/15103)
[Closes#1661](https://github.com/twentyhq/core-team-issues/issues/1661)
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
## Problem
CI workflow started timing out on October 14, 2025 after commit
`d750df7fff` removed the trailing newline from `.env.example`.
## Root Cause
When `.env.example` lacks a trailing newline:
```bash
# Last line without newline
# CLICKHOUSE_URL=...twenty
```
And CI runs:
```bash
echo "NODE_PORT=3002" >> .env
```
Result:
```bash
# CLICKHOUSE_URL=...twentyNODE_PORT=3002 ← Commented out!
```
Server starts on default port 3000 instead of 3002, health check fails.
## Fix
1. **Restore trailing newline** to `.env.example`
2. **Make all CI `.env` operations robust** by adding `echo "" >> .env`
before appending
3. **Simplified `set_env_var`** function to always add newline first
Now works regardless of whether template files have trailing newlines.
## Files Changed
- 6 CI workflow files
- 1 .env.example file
## Summary
Clean up and consolidate formatting configuration across the monorepo.
## Changes
### 1. Remove Redundant Configs
- ❌ Delete `packages/twenty-server/.prettierrc` (had invalid
`brakeBeforeElse` typo)
- ❌ Delete `packages/twenty-zapier/.prettierrc`
- ✅ Use root `.prettierrc` only (Prettier searches up directory tree
automatically)
### 2. Improve Root Prettier Config
- Change `endOfLine: 'auto'` → `'lf'` for consistent Unix line endings
across all OSes
### 3. Enhance VSCode Settings
- `files.eol: 'auto'` → `'\\n'` (consistent with Prettier)
- Add `files.insertFinalNewline: true` (explicit editor behavior)
- Add `files.trimTrailingWhitespace: true` (cleaner files)
### 4. Add `.gitattributes`
- Enforce LF line endings at Git level
- Prevents `core.autocrlf` from converting based on contributor's OS
- Mark patch files as binary (they have mixed line endings by design)
- Explicitly define binary file types
### 5. Fix Line Endings
- Convert 3 selectable-list state files from CRLF → LF
- These were the only source files with Windows line endings
## Why These Changes Matter
**Before:**
- 3 different Prettier configs (inconsistent, one had typo)
- Mixed CRLF/LF depending on contributor's OS
- No Git-level enforcement
**After:**
- Single source of truth for formatting
- All files use LF (Unix standard)
- Git enforces line endings regardless of OS
- Prettier warning about invalid option removed
## Result
- ✅ Single `.prettierrc` config
- ✅ Consistent LF line endings enforced by Git
- ✅ Better VSCode defaults
- ✅ No more CRLF files sneaking in from Windows contributors
## Summary
**Step 1 of 2:** Implements the "acting on behalf of user" concept for
workflows and agents to prevent permission escalation and maintain
proper audit trails.
## Problem
Previously, workflows and agents would bypass permissions regardless of
who initiated them, allowing users to escalate their privileges by
triggering workflows that performed actions they couldn't do directly.
## Solution
### For Workflows
Introduced `WorkflowExecutionContext` service that determines execution
mode:
- **Manual triggers/test button**: Uses user's roleId for permissions,
user's identity for `createdBy`
- **Automated triggers** (cron, database events, webhooks): Bypasses
permissions, uses workflow identity
### For Agents
**In Chat:**
- Always act on behalf of the user
- Use user's roleId for permission checks
- Use user's identity for `createdBy`
# Step 1 vs Step 2
### ✅ Step 1 (This PR): Acting on Behalf Concept
- Introduced `isActingOnBehalfOfUser` boolean concept
- Single roleId used for permission checks (user's OR system bypass)
- `createdBy` field properly attributes actions to initiator
- Prevents permission escalation in user-initiated flows
### 🔜 Step 2 (Future): Multi-Role Permission Support
- Support role intersection: `{ intersection: ['roleA', 'roleB'] }`
- Support role union: `{ union: ['roleA', 'roleB', 'roleC'] }`
- Enable user+agent collaboration scenarios
- Update `WorkspaceEntityManager` and `WorkspaceDatasource` to handle
multiple roleIds
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
## Context
- All flatEntity should extend SyncableEntity
- SyncableEntity should now have applicationId and application relation
- Fix syncApp deletion, should now properly use migration v2 to delete
syncable entities
# Introduction
### Summary
Implements side effect handling for `ViewGroup` and `ViewFilters` when
field metadata is updated in the v2 architecture. This ensures that
view-related records are properly maintained when enum field options are
modified, deleted, or created.
### Side effects
- **Side Effect System**: Added side effect handling for field metadata
updates that manages related view groups and view filters
- **Enum Field Updates**: When enum field options are modified, the
system now:
- **View Groups**: Creates new groups for added options, updates
existing groups for modified options, and deletes groups for removed
options
- **View Filters**: Updates filter values to reflect option changes and
removes filters that reference deleted options
### Enum runner fix
Update now works for both atomic enum and array enum ( multi select for
instance )
### Compute flat entity maps from to
Standardized this method usage across v2 services
Next step is to require dependencies dynamically
## Conclusion
closes https://github.com/twentyhq/core-team-issues/issues/1649
Improvements to database seeding performance and developer experience.
**Changes:**
1. **Attachment seeding**: Add sample files (PDF, XLSX, PPTX, PNG, ZIP)
to dev seeds with proper file storage
2. **Seeding parallelization**: Process entities within batches in
parallel while respecting dependencies
3. **ORM query logging**: Replace manual logger toggling with
`ORM_QUERY_LOGGING` env var
- Values: `disabled` (default), `server-only` (for local dev), `always`
- Configured once in `core.datasource.ts`, removed from all seeder
services
**For .env:**
```bash
ORM_QUERY_LOGGING=server-only
```
Net result: Faster seeding, cleaner code (-68 lines), better local dev
experience.
I had an issue with invalid UUIDs, and I think it would be easier to
find the offending one if the UUID were included in the error message.
This way a database dump can be easily searched.
---------
Co-authored-by: prastoin <paul@twenty.com>
This commit fixes a PostgresException error that occurred when
processing timeline activities with empty workspaceMemberId values.
workspaceMemberId appears to be optional but we set it to an empty
string when it is not present. This subsequently causes issues in the
postgres query.
This change fixes the root cause. Empty string handling is added to the
postgres query also for situations where bad data has already entered
the db
Example error
```
[Nest] 34 - 10/13/2025, 9:09:55 PM LOG [BullMQDriver] Job 2274 with name MessageParticipantMatchParticipantJob processed on queue messaging-queue
query failed: SELECT "timelineActivity"."happensAt" AS "timelineActivity_happensAt", "timelineActivity"."name" AS "timelineActivity_name", "timelineActivity"."properties" AS "timelineActivity_properties", "timelineActivity"."linkedRecordCachedName" AS "timelineActivity_linkedRecordCachedName", "timelineActivity"."linkedRecordId" AS "timelineActivity_linkedRecordId", "timelineActivity"."linkedObjectMetadataId" AS "timelineActivity_linkedObjectMetadataId", "timelineActivity"."id" AS "timelineActivity_id", "timelineActivity"."createdAt" AS "timelineActivity_createdAt", "timelineActivity"."updatedAt" AS "timelineActivity_updatedAt", "timelineActivity"."deletedAt" AS "timelineActivity_deletedAt", "timelineActivity"."workspaceMemberId" AS "timelineActivity_workspaceMemberId", "timelineActivity"."personId" AS "timelineActivity_personId", "timelineActivity"."companyId" AS "timelineActivity_companyId", "timelineActivity"."opportunityId" AS "timelineActivity_opportunityId", "timelineActivity"."noteId" AS "timelineActivity_noteId", "timelineActivity"."taskId" AS "timelineActivity_taskId", "timelineActivity"."workflowId" AS "timelineActivity_workflowId", "timelineActivity"."workflowVersionId" AS "timelineActivity_workflowVersionId", "timelineActivity"."workflowRunId" AS "timelineActivity_workflowRunId" FROM "workspace_8h07bh3zq5pjg65lx9qbxbgg0"."timelineActivity" "timelineActivity" WHERE ( (("timelineActivity"."noteId" IN ($1, $2)) AND ("timelineActivity"."name" IN ($3, $4)) AND ("timelineActivity"."workspaceMemberId" IN ($5, $6)) AND ("timelineActivity"."createdAt" > $7)) ) AND ( "timelineActivity"."deletedAt" IS NULL ) ORDER BY "timelineActivity"."createdAt" DESC LIMIT 1 -- PARAMETERS: [null,"aa6f2e7f-16b1-447f-9988-0ba359358609","linked-note.created","note.created","",null,"2025-10-13T20:59:55.267Z"]
error: error: invalid input syntax for type uuid: ""
/app/packages/twenty-server/dist/src/engine/twenty-orm/error-handling/compute-twenty-orm-exception.js:36
throw new _postgresexception.PostgresException(error.message, errorCode);
^
PostgresException [Error]: invalid input syntax for type uuid: ""
at computeTwentyORMException (/app/packages/twenty-server/dist/src/engine/twenty-orm/error-handling/compute-twenty-orm-exception.js:36:19)
at WorkspaceSelectQueryBuilder.getMany (/app/packages/twenty-server/dist/src/engine/twenty-orm/repository/workspace-select-query-builder.js:56:76)
at process.processTicksAndRejections (node:internal/process/task_queues:105:5)
at async WorkspaceRepository.find (/app/packages/twenty-server/dist/src/engine/twenty-orm/repository/workspace.repository.js:33:24)
at async TimelineActivityRepository.findRecentTimelineActivities (/app/packages/twenty-server/dist/src/modules/timeline/repositories/timeline-activity.repository.js:68:16)
at async TimelineActivityRepository.upsertTimelineActivities (/app/packages/twenty-server/dist/src/modules/timeline/repositories/timeline-activity.repository.js:27:42) {
code: '22P02'
}
```
# Introduction
Fixed the build dependency leading to twenty-server start failing before
building,
Removed redundant steps
Make everything run on test db
This migration updates the filter operand values of the workflowVersion
and workflowRuns in order to capitalize them as they should be (the
product works with both deprecated camel case and capitalized). But that
may involve thousands of workflowRuns!
Let's update the command to only update workflowVersion, and update the
cleanWorkflowRuns job to remove workflow runs that are more than 14 days
old.
This way after the command is run, all new workflow runs will have the
new value for the filter operand, and after fourteen days there will be
no trace of the workflow runs with the deprecated filter operand.
# Introduction
Defining very first basis of the twenty-cli e2e testing env.
Dynamically generating tests cases based on a list of applications names
that will be matched to stored twenty-apps and run install delete and
reinstall with their configuration on the same instance
We could use a glob pattern with a specific e2e configuration in every
apps but right now overkill
## Notes
- We should define typescript path aliasing to ease import devxp
- parse the config using a zod object
## Some vision on test granularity
Right now we only check that the server sent back success or failure on
below operation. In the future the synchronize will return a report of
what has been installed per entity exactly. We will be able to snapshot
everything in order to detect regressions
We should also be testing the cli directly for init and other stuff in
the end, we could get some inspiration from what's done in preconstruct
e2e tests with an on heap virtual file system
## Conclusion
Any suggestions are more than welcomed !
close https://github.com/twentyhq/core-team-issues/issues/1721
## Goal
- inferDeletionFromMissingEntities is now a map instead of a single
bool, allowing us to parameterise it based on the entity we want to
compare
- Adding index creation/update when field isUnique is set to true and
index deletion when isUnique is false (for now)
close https://github.com/twentyhq/core-team-issues/issues/1346
### Summary
Split BAR into VERTICAL_BAR and HORIZONTAL_BAR as separate chart types.
### The Problem
Initially wanted a simple vertical/horizontal toggle for bar charts, but
ran into a GraphQL union type
constraint: union types can't have the same field name with different
nullability.
- Vertical bars need groupByFieldMetadataIdX as required (categories on
X)
- Horizontal bars need groupByFieldMetadataIdY as required (categories
on Y)
- GraphQL schema generation fails with this setup
### The Solution
Use semantic primaryAxis and secondaryAxis naming that's
orientation-agnostic:
- primaryAxisGroupByFieldMetadataId = main grouping field (e.g.,
"Company Name")
- secondaryAxisGroupByFieldMetadataId = optional secondary grouping
(e.g., "Stage")
These fields have consistent meaning regardless of orientation. The
visual mapping happens at the UI layer:
- Vertical bars: primary data renders on X-axis, secondary on Y-axis
- Horizontal bars: primary data renders on Y-axis, secondary on X-axis
Both chart types share the same DTO structure with consistent
nullability.
### What Changed
- Split GraphType.BAR → VERTICAL_BAR | HORIZONTAL_BAR
- Renamed fields: primaryAxisGroupByFieldMetadataId,
secondaryAxisGroupByFieldMetadataId (+ subfield
variants)
- useChartSettingsValues(): Maps semantic fields to setting values (no
swapping)
- getBarChartSettings(): Dynamically arranges settings panel based on
orientation
- transformGroupByDataToBarChartData(): Maps semantic fields to Nivo's
layout prop
video QA
https://github.com/user-attachments/assets/479061b5-712e-4ca6-9858-95273d1f16c1
Closes#14726
### Added
- `trashRetentionDays` field to workspace entity (default: 14 days)
- Automated trash cleanup using BullMQ jobs
- Daily cron (00:10 UTC) that enqueues cleanup jobs for all active
workspaces
- Per-workspace limit: 100k records deleted per day
- Calendar-based retention: records deleted on day X are cleaned up X+14
days later (at midnight UTC boundaries)
### Architecture
- **Cron (WorkspaceTrashCleanupCronJob):** Runs daily, enqueues jobs in
parallel for all workspaces
- **Job (WorkspaceTrashCleanupJob):** Processes individual workspace
cleanup
- **Service (WorkspaceTrashCleanupService):** Discovers tables with
`deletedAt`, deletes old records with quota enforcement
- **Command:** `npx nx run twenty-server:command
cron:workspace:cleanup-trash` to register the cron
### Testing
- Unit tests for service with 100% coverage of public API
- Tested quota enforcement, error handling, and edge cases
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
## Problem
When trying to signup on localhost:3001 with a new random email, users
were getting an error message saying 'User already exists', even though
they were new users.
## Root Cause
The `signUpWithoutWorkspace` method in `sign-in-up.service.ts` had
inverted logic when checking if a user exists. It was using
`findUserByEmailOrThrow` which:
- **Returns the user** if found
- **Throws the provided error** if NOT found
This caused the opposite behavior:
- ❌ **New user (doesn't exist)**: Threw 'User already exists' error
- ❌ **Existing user**: Continued to create duplicate user
## Solution
Changed to use `findUserByEmail` and explicitly check if the user exists
before throwing the appropriate error:
```typescript
const existingUser = await this.userService.findUserByEmail(newUserParams.email);
if (existingUser) {
throw new AuthException(
'User already exist',
AuthExceptionCode.USER_ALREADY_EXIST,
{ userFriendlyMessage: msg`User already exists` },
);
}
```
This matches the correct pattern already used in `signUpInWorkspace`
(line 431-441 in auth.resolver.ts).
## Changes
- Fixed inverted logic in `signUpWithoutWorkspace` method
- Now correctly validates that user does NOT exist before creating new
user
- Matches the pattern used in `signUpInWorkspace`
## Testing
The fix corrects the logic so that:
- ✅ New users can sign up successfully
- ✅ Existing users get the correct 'User already exists' error
I did an acceptable design for empty node and iterators for release:
- use array field for iterator node. Added an util to stringify arrays
for backward compatibily
- remove icon for empty node
- allow to select a node on empty node selection
https://github.com/user-attachments/assets/b00037a8-aa1d-4784-b973-05973649b46e
# Introduction
Adding view-group to core engine v2
Following https://github.com/twentyhq/twenty/pull/15010 ( same pattern )
## What's done
- Created flat-view-group
- flat view group runner
- flat view group builder
- create view group service v2 and input transpilers
- refactor the existing view group resolver to fix standard (
BREAKING_CHANGE on graphql api update especially ) REST stays the same
- refactored the front to consume the mutations autogenerated
close https://github.com/twentyhq/core-team-issues/issues/1665
To avoid huge workflows to block the worker, we will enqueue a new job
every 20 steps.
This could be more than 20 if there are branches but I think this is
fine, the goal is only to have a limit set.
Also cleaning a bit the code to mark running steps as failed when
workflow fails.
I tested it on a huge workflow:
https://github.com/user-attachments/assets/d7b8e345-d1a1-4467-96fd-92117b500120
