Applications

Developer toolchain / Agents

Foundry now makes it easier to build, configure, and ship pro-code agents. Agents combine a large language model with your Foundry data and tools. They can read and write Ontology data, fix failing builds, or migrate legacy systems into Foundry. Agents now authenticate against the Ontology SDK (OSDK), Ontology MCP (OMCP), and Palantir MCP with scoped permissions out of the box, so you no longer pass a client ID and secret to call tools. After you publish an agent, you can call it from Workshop or OSDK with no additional configuration.

Features

Ontology building / Automate

Automate now supports Global Branching. You can modify automations on a branch and test them end-to-end before merging your changes into the main branch. Branching lets you iterate on automation logic without disrupting live workflows or the people who depend on them. Automate supports the full set of branching features: modifications, environment isolation, protection, rebasing, and approvals.

Features

Data connectivity & integration / SQL Studio

SQL Studio now supports autocomplete for branch name prefixes in queries. When writing queries that reference a specific branch, the editor will suggest matching branch names as you type.

Enhancements

Analytics / Quiver

Quiver's right-side configuration panel now fully supports dark mode. When dark mode is enabled, the editor panel, dialogs, overlays, and code viewers render with proper dark styling. This includes editors across all card categories.

Data connectivity & integration / SQL Studio

SQL Studio now displays all data sources referenced in your SQL query in the sidebar. This provides a quick overview of the datasets and object types being used without needing to inspect the query text directly.

Applications

AI Platform (AIP) / Language Model Service

Claude Sonnet 5 is now available in AIP for eligible commercial and US government enrollments. Review the announcement to learn more about the model's availability on your enrollment.

Features

Model connectivity & development / Modeling

Model evaluations is a new Python API for capturing how a model version performs against test data and visualizing the results directly on the model page. An evaluation is a collection of metrics, images, plots, and tables that you define and log yourself, allowing you to compare model performance across versions and over time.

Enhancements

Product delivery / Marketplace

You can now package and deploy versioned object sets built using Object Explorer or Insight in Marketplace.

Features

Analytics / Insight

Insight workbooks now autosave your in-progress work. As you edit a workbook, unsaved tabs, layouts, filters, and object set selections are automatically preserved and restored after a refresh or navigation. You can also copy the workbook URL to share the current state of your analysis with others.

Enhancements

Analytics / Notepad

You can now insert images into Notepad documents directly using the / keyboard shortcut. Select the Image option to open your computer's file picker and upload an image without needing to drag and drop or paste from the clipboard.

Data connectivity & integration / Pipeline Builder

Pipeline Builder now supports fallback models in LLM transforms. You can specify a prioritized list of models so that if the primary model fails, subsequent models are automatically attempted in order. Fallback occurs only when the current model encounters a non-retryable error and no successful response has yet been generated. When a fallback is triggered, a debug message notifies you of the model switch.

Data connectivity & integration / SQL Studio

Object types referenced in a SQL query now automatically appear in the SQL Studio sidebar. This allows users to view the object types being queried without navigating away from the editor.

Ontology building / Workflow Lineage

When viewing a Workflow Lineage graph that includes object types, action types, or interfaces from multiple ontologies, all color modes (such as locations, last updated, permissions, and monitors) now apply to those cross-ontology resources. Previously, resources from a different ontology than the currently selected one appeared gray regardless of the active color mode, making it difficult to assess their status. Cross-ontology resources still display with a distinct gray header in the default color mode to differentiate them from resources in the current ontology.

Ontology building / Workflow Lineage

When you right-click an object type, action type, function, AIP Logic, or Automate node in Workflow Lineage, the context menu now includes two new monitoring options: Create monitor rule and View all alerts. Selecting Create monitor rule opens a new tab where you can configure a new monitor rule for that resource, while View all alerts opens a new tab showing all alerts associated with the resource's existing monitors. These options appear only for resources that support monitoring.