> For the complete documentation index, see [llms.txt](https://cleyrop.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://cleyrop.gitbook.io/docs/documentation-fr-en/support-and-resources/changelog-and-release-notes/release-note-v4.3.md).

# Release Note v4.3

### 📢 Announcements

Cleyrop version 4.3 introduces the **AI Contexts**, a new capability enabling the creation and sharing of specialized AI assistants within projects. This version also enhances the **low-code transformations** with 4 new operations (Union, Text, Concatenate, Top N), improves the **dataset management** with refresh from local import and the addition of column descriptions, and strengthens the**API** with the exposure of extended metadata. Finally, significant improvements concern the**integrated SQL exploration** for dataflows and the **Snowflake connector** which now supports views and queryable objects.

***

### 🚀 Feature additions

#### AI Contexts in AI Studio

**Create and manage reusable AI capabilities**

* Creation of custom AI contexts with structured prompt, Agent Tools (MCP) and knowledge scope
* Lifecycle management: Test, Publish, Edit, Delete, Share
* Use AI contexts directly in the Global Assistant
* Dedicated conversation history per AI context
* Configuration of user environment variables for MCPs requiring personal credentials

#### Low-Code transformations

**4 new no-code transformations**

* **Union** : Vertically combine multiple datasets with similar structure with schema handling and deduplication option
* **Text** : Apply string manipulation operations (LEFT, RIGHT, TRIM, UPPER, LOWER, REPLACE) on one or more columns
* **Concatenate** : Merge several text columns into a new column with customizable separator and NULL value handling
* **Top N** : Extract the first or last N rows according to a sort, with the option to partition by group and add a rank column

#### Dataset Management

**Enhancement of management capabilities**

* **Refresh datasets from local import** with schema validation, column drift handling and mode selection (Append/Replace)
* Addition of **descriptions** to columns to document business meaning and improve data governance
* Support for views and queryable objects (standard, materialized, secure, dynamic views and external tables) in SQL-type data sources (Snowflake, PostgreSQL...)
* Detailed display of ingestion logs for failed datasets or datasets requiring attention (error limited to the first detected occurrence)
* SQL exploration of intermediate results within a dataflow

#### API - Extended Metadata

**Enhancement of the dataset metadata endpoint**

* Addition of the dataset description
* Addition of labels (tags) associated with the dataset
* Available via the endpoint `/v1/datasets/{id}/metadata`

#### Cleyrop S3 Library

**New functions for advanced interactions**

* Bucket and object existence checks
* Metadata retrieval (LastModified, ETag, Size)
* Object listing pagination
* Multipart upload support

#### Data quality

**360° view of quality rules**

* Consolidated view of dataset rule results in the "My subscriptions" tab

#### The Factory of Agents

**Deployment management**

* Ability to cancel an ongoing instance deployment
* Configuration of deployment timeout by instance (default value: 9 minutes)
* Addition of public Git repositories without the need for a deploy key

***

### 🐞 Bugfix

* Fix for the "Ignore schema types" button, which did not reset column types correctly
* Fix for uppercase path support during image build
* Fix for handling recurrence changes on datasets