---
title: "28th of August 2026"
slug: "28th-of-august-2026"
status: "new"
updated: 2026-09-03T06:27:47Z
published: 2026-09-03T06:27:47Z
canonical: "knowledge.curiositysoftware.ie/28th-of-august-2026"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://knowledge.curiositysoftware.ie/llms.txt
> Use this file to discover all available pages before exploring further.

# 28th of August 2026

### New Features

- **Pipeline export/import between environments** — pipelines (including variables, model logic, and activity/runspec mappings) can now be exported to a file and re-imported into another environment, either interactively or via an automated VIP process, making it easier to ship pipeline accelerators between environments.
- **Machine Learning Rule Set Accelerators** — a new action for creating ML Rule Set accelerators, plus supporting library activities.
- **Feature Association** — a new object type, working alongside Foreign Keys, for linking related columns across tables.
- **New activity actions** — Transform Rule Set Accelerators, Create SQL / Preview row count (Subset Data Activity), and Analyze Data are now available as actions.
- **Create Group (TDM library)** — table groups can now be created directly from the TDM library.

### Enhancements

- Cross-definition foreign keys are now shown in ER diagrams, so relationships to tables in other data sources are visible.
- Spot and ER diagrams can now be embedded in iframes, matching existing submit-form support — useful for surfacing diagrams in Confluence/SharePoint.
- Oracle masking can now use ROWID as the unique identifier for tables without a primary or unique key.
- Pipeline variables can now be sorted (by index, name, description, UI type, or group) in the variable screen.
- New Excel template keywords for pulling a variable's final value.
- Data Generation: table names in the bulk copy activity can be changed via parameter; a custom generation function can be specified for disabled columns.
- Scan Activity and Subset Data Activity gained several new actions and dropdown options.
- Various fixes to Export/Import of Ruleset details (scoped variables, table/column ordering, pre/post processes).
- Job engine status now catches cases like repeated server restarts.
- MCP: JSON passed to AI now includes dropdowns; MCP TDM can now handle table groups; pipeline jobs can be run through MCP as per the normal submission process.

### Bug Fixes

- Fixed rule handling over subflows and parallel gateways.
- Fixed duplicate/inconsistent variable naming when the same subflow is used twice, affecting Preview Test Data and Excel export
- Fixed exported test steps losing subflow variable values after deleting suite paths.
- Fixed data tables incorrectly appearing as test steps in exports.
- Fixed Master Model cloning timing out for models with a large number of Business Components.
- Fixed Custom Field and Test Data not coming through correctly on test case export to Zephyr.
- Fixed Message Management submit forms silently using incorrect parameters when no template is attached.
- Fixed several UI/navigation issues: ribbon not working from pre/post processes, negative block location in pipelines, ARD imported subflows rendering stacked, job link lost when triggering a pipeline from a submission form.
- Fixed Modeller creating duplicate Test ID links in Jira/Xray on every model update and export.
- Fixed AI Scanning Result export producing `[object,object]` instead of readable text.
- Fixed ML Data Generation failing when a table is present in a foreign key but not in the rule set, and when deleting a table from a generation rule set.
- Fixed ML Training Rule Set configuration edits not saving.