> For the complete documentation index, see [llms.txt](https://docs.datasaur.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.datasaur.ai/compatibility-and-updates/release-notes/version-6/6.163.0.md).

# 6.163.0

June 11, 2026

## **What's New? 💡** <a href="#whats-new" id="whats-new"></a>

### **Data Studio** <a href="#data-studio" id="data-studio"></a>

* Adding project kinds to ongoing projects — Users can now add additional project kinds to an existing project, enabling workflows that combine multiple labeling types within a single project.
* Audio Labeling:
  * Users can now set a custom skip forward and backward duration in the audio controls, with support for up to one decimal place. The preference is saved across sessions.
  * Multi-channel audio projects now display audio tracks in a side-by-side layout by default, allowing users to view and label multiple channels simultaneously.
* Advanced Search — Clicking a search condition chip now directly opens the Advanced Search logic builder, reducing the steps needed to edit existing search conditions.
* Review Extension — The "Mixed value" status label in the Document status has been renamed to "Mixed statuses across labelers" to more clearly indicate when annotators have different document completion statuses.
* Bounding Box Labels Extension — The box and arrow count display has been updated to show total counts and hidden counts separately for bounding boxes and arrows.
* Arrow Labeling Agent — Users can now configure and run an Arrow Labeling Agent for Span-based projects, enabling automated arrow label prediction powered by LLM Labs. Arrow agent tasks can be set up in the Project Creation Wizard and adjusted from within the label editor.
* Row-based Labeling Agent — Users can now configure individual tasks for each agent within a Row-based Labeling Agent assignment.
* Labeling Agent — When an LLM Labs credit-limit error occurs during a labeling agent job, the job now halts after the first failed inference to prevent unnecessary quota consumption.

## **Bug Fixes 🐞**

### **Data Studio** <a href="#data-studio" id="data-studio"></a>

* Fixed an issue in Row Labeling projects where labeling agent progress was not displayed in the Members column on the Projects page.
* Fixed an issue in Span Labeling where changing the "Show labels from inactive label set in Review mode" setting did not take effect immediately and could cause duplicate labels when switching between label sets.
* Fixed an issue in Span Labeling projects where the "Labeled by" field in the label tooltip was empty for labels predicted using ML-assisted labeling in Reviewer mode.
