Reactor Data Monthly Release Notes Digest: July 2025 - Knowledge Base

Even as summer heats up, the Reactor Data team has been hard at work this past month, delivering a wave of new features and key improvements across the platform. Dive into the digest below, grouped by UI theme, to see how we're making your Reactor experience even better.

Electron

Electron's "Act/Plan" Toggle: We have added an "Act/Plan" toggle in Electron, giving you more control over how Electron applies changes to your mappings. You can now choose whether Electron auto-applies changes or presents a plan for your review first. Read more about this feature here.
Reliable Electron Prompt Execution: Electron is now more reliable at successfully executing mapping changes based on your prompts.
Electron Upgrade for Metadata Fields: Electron is now smarter! It automatically uses pre-defined mappings and descriptions for Reactor-metadata-specific field names (such as scid, reactor_is_redacted, loaded_at, etc.), ensuring greater consistency with no extra effort from your side. You can always overwrite Electron's suggested mappings, but now you don't need to cross-reference our landing table best practices page to remember how to map any of these key fields.

Sources

Expanded Data Previews: The Data Preview panel for new data sources in Sources now displays up to 100 sample records (previously 25), providing a much more comprehensive initial look at your data.

Mappings

Accurate Source Link Displays: We've corrected an issue where Source Link lines in the Mappings UI weren't updating after a Source Link was deleted. The UI now accurately reflects your active Source Links.
"Cancel" Button Refinement: We have added a "Cancel" button to the Expression Editor. When selected, Reactor will discard changes to the specific field you're working on and revert to it's last-deployed state.
Streamlined Sandbox Input Deletion: You can now easily delete inputs from a Model in the Mappings Sandbox using the trash icon.
Input Label Name Error Fixed: We've resolved the "Label name is already in use" error that sometimes occurred when adding new inputs to a model on the Mappings page.
Improved Error Tab in Mapping Results: The Errors tab in the Mapping Results panel has been enhanced with sorting and filtering capabilities, a scroll bar for long lists of errors, and the display of the full field path for errors within nested fields (e.g., object.field instead of just field).
More Specific Error Messages for Complementary Inputs: When evaluating a model that includes both primary and complementary inputs, you will now receive a more specific error message if you attempt to evaluate the model against a complementary input, instructing you to select a primary input source instead.
Smoother Configuration Saves: We fixed a bug where saving a model configuration in Production would close it. Now, your model stays open after saving, allowing you to continue working uninterrupted.
No Refresh for Sandbox Saves: We've updated the underlying procedure for saving a model configuration so that your screen will no longer refresh after saving a Production model to Sandbox, removing an unnecessary disruption to your workflow.
Infinite Request Loop Resolved: A bug that caused an infinite request loop when trying to add a new model has been resolved.

Destinations

Unstuck the Add Sandbox Target workflow: We've resolved an issue that caused Reactor to hang when adding sandbox output targets on the Destinations page, ensuring a smoother deployment process.