twenty-cli serverless triggers follow up. Fixes:
- eventName don't support wildcard
- universalIdentifier not used to create or update trigger : update does
not work properly (does deletion then creation)
- add a base project in twenty-cli that is copied when creating a new
app
In this PR https://github.com/twentyhq/twenty/pull/14785 we deprecated
viewFilterOperand (from camelCase to capital snakeCase values), but we
had not migrated the existing workflow steps values. Hence they cannot
be executed!
Let's fix that by dealing with both operands, new and deprecated, to
mitigate the issue. Then we will add a command to migrate the values.
Streamline service definitions in the Makefile by ensuring the
`ensure-docker-network` dependency is called directly in each service
target instead of manual invocation. This change simplifies the process
of creating Docker services.
Closes#14715
Tested :
- server level ok
- workspace level ok
- server level without canImpersonate or without allow impersonation -
fail ok
- workspace level without permission - fail ok
- add serverlessFunction schema in twenty-cli
- add trigger schema in twenty-cli
- update serverless function code save and get
- sync serverless function
# Calendar View: Add "+" Button to Calendar Day Cards
## Description
This PR implements the "+" button on calendar day cards in the Month
view, allowing users to create new records directly from the calendar.
- The "+" button appears **on hover** only.
- Newly created records include the **day representing the card** in the
payload (UTC, with only the date part).
- Respects **object permissions** and **soft delete filters**.
- Automatically integrates with **active filters** used in Table and
Kanban views.
## Implementation Notes
- `RecordCalendarAddNew` component handles button visibility and record
creation.
- UTC date is passed to the create function to ensure proper timezone
handling.
- No unit tests added yet (need senior guidance).
---
[Screencast from 2025-09-24
20-10-18.webm](https://github.com/user-attachments/assets/bd117b43-6c1b-4267-874f-b36844733d9e)
### Video ->
## Related Issues
Fixes: https://github.com/twentyhq/core-team-issues/issues/1552
---------
Co-authored-by: Weiko <corentin@twenty.com>
Legacy from early v2 implementation, we don't want to store relation ids
in the cache as the new implementation handle that with flat entities
properly already.
This causes some issues because when we return the updated view, it
contains an array of viewField objects with only the id in it so gql
complains because some fields are non-nullable in the output and the
resolver can't resolve viewField directly because viewField is in the
output (not complete though) already so it just ignores it.
Following https://github.com/twentyhq/twenty/pull/14762
In this PR
- moved logic around any field filter formatting to twenty-shared
- added any field filter to the filters applied to groupBy when viewId
is defined
- added integration test
In the BE we have stored filter operand as IS_EMPTY, IS_NOT_EMPTY, etc.
For some reason in the FE we were manipulating IsEmpty, IsNotEmpty, etc.
(maybe because they were used before in Views before they were moved to
core)
So we were converting the operands in the FE from IS to Is
(convertViewFilterOperandFromCore) to read and manipulate viewFilters,
and then back to BE version to send mutations etc., from Is to IS
(convertViewFilterOperandToCore).
The migration is now over, so we can remove and simplify that code.
(In the 1-5:migrate-views-to-core command we still do the migration from
Is format to IS format.)
When all items have been processed, iterator should:
- store last iteration in history
- return the final result saying iteration have been successful. So it
gets stored by the executor
It was not doing the first step, so we were losing the last iteration.
Splitting into separated function + adding tests
Fixes https://github.com/twentyhq/twenty/issues/14569
Some extensions, such as SimilarWeb, are incompatible with twenty: they
trigger intempestive "AbortErrors" converted into error snackbar. While
twenty remains usable it is very annoying.
I did not take time to deeply investigate which requests are
problematic. According to ClaudeAI it may be because twenty makes a lot
of queries to the same endpoint /graphql, a pattern that SimilarWeb may
interpret as tracking which they try to intercept.
In this PR, we finish the work started in [this
PR](https://github.com/twentyhq/twenty/pull/13080) to silent harmless
AbortErrors.
