Version 2026.38
24 changed1 retiring3 fixed
Amazon Advertising¶
With this release, proactive warning notifications and graceful handling of expired refresh tokens have been implemented for the Amazon Advertising connector. This change addresses Amazon’s new 365-day refresh token expiration policy, ensuring uninterrupted data collection and providing clear guidance for re-authorization. For more information, see Amazon Advertising: Setup guide and the announcement.
AppLovin¶
With this release, obsolete or unused datastream templates have been removed from the AppLovin connector, streamlining the selection process for users.
Facebook Ads¶
With this release, a new System User Access Token authorization method has been added to the Facebook Ads connector, providing a more secure and stable alternative to personal OAuth2 tokens for long-term data collection. Additionally, a warning system now alerts you if your ad accounts are not opted in for specific Insights API breakdowns, such as impression_device, hourly_stats_aggregated_by_audience_time_zone, and frequency_value, helping to prevent silent data loss. For more information, see Facebook Ads: Setup guide.
Google Search Ads 360¶
With this release, the Google Search Ads 360 (beta) connector now includes a richer set of asset and asset set details within the Campaign and Ad Group report types. New fields such as Asset Id, Asset Type, Asset Final Urls, and Asset Set Id are now available, providing more comprehensive data for analyzing ad performance.
Innovid¶
With this release, new fields version_id and version_name have been added to the Innovid connector for the Campaign Delivery (Custom Fields) report type, providing more granular insights into campaign delivery data.
Instagram Business¶
With this release, the Instagram Business connector’s Comments report type now includes the author’s username and id for each reply within the Replies content. This enables accurate community management tracking and sentiment analysis.
Integral Ad Science¶
With this release, new fields Advertiser and Advertiser ID have been added to the Integral Ad Science (IAS) Performance report type, enabling more detailed analysis and reporting of advertising data.
LinkedIn Ads¶
With this release, the LinkedIn Ads connector has been improved by adding validation for empty Ad accounts in the Include/Exclude filter for the Metadata report type. This prevents misconfigurations that could lead to incomplete or incorrect data. For more information, see LinkedIn Ads: Setup guide.
MySQL destination¶
With this release, the MySQL destination now supports SSL-encrypted connections, including CA certificate upload and validation. This allows users to establish secure connections, meeting security requirements and protecting sensitive data in transit.
Reddit Ads¶
With this release, the Reddit Ads connector has been improved with a new Ad Status filter in the Analytics report, allowing users to include “DELETED” ad statuses. This provides a more comprehensive view of ad performance, including historical data for deleted ads. For more information, see Reddit Ads: Setup guide.
SEMRush¶
With this release, the SEMRush connector has been updated for Domain Organic Search Keywords and URL Organic Search Keywords report types. The deprecated Fl SERP Features column has been replaced with Fp (Positional SERP Features), and Fk (Keywords SERP Features) has been added as a new selectable field. This provides users with updated and more granular SERP Features data, improving their ability to analyze search engine results.
Sprout Social¶
With this release, a new Cases report type has been added to the Sprout Social connector. This enables comprehensive reporting on customer care interactions and related metrics. For more information, see Sprout Social: Setup guide.
TikTok Shop¶
With this release, the TikTok Shop connector has been improved with a new Order status filter for Order detail reports. Additionally, Order return reports now feature selectable metric field sets and improved flattening of array fields into separate output columns, providing more structured and usable information. For more information, see TikTok Shop: Setup guide.
Set custom schedule dialog¶
With this release, custom schedule entries for a datastream are now managed together in a single Set custom schedule dialog: add, edit, or remove any number of entries and save them all at once, instead of configuring one scheduled fetch at a time. For more information, see Manual, Smart, and Custom fetches.
Collapsible Custom Script instructions¶
With this release, Custom Script instructions can be collapsed to a single header row, making transformations with many instructions easier to navigate and reorder. Use Collapse all / Expand all to toggle every instruction at once. Instructions with validation or preview errors are highlighted, show an error count in the header, and expand automatically. For more information, see Using custom script transformations.
System Environment page¶
With this release, the System Environment page is now available to workspace administrators. It displays your instance’s version, URL, IP address, and IAM Role ARN, so you can retrieve the exact role ARN needed for your AWS trust policy without contacting Adverity Support. For more information, see Viewing system environment details.
Datastream copying API¶
With this release, copying a datastream into a different workspace (by passing that workspace’s ID as stack) with POST /api/datastreams/{id}/copy/ follows the same rules as the Clone wizard. Configuration, schedules, and column mappings are carried over to the copy. The authorization is reused if the target workspace can use it, and cloned into the target workspace otherwise. Destination bindings, enrichments, and monitors of the source datastream are no longer carried over. Use the new optional destinations, enrichments, and authorization_id request fields to choose what the copy uses. The response includes a new report field that lists what was reused, cloned, attached, and dropped. Errors now return HTTP 400 instead of HTTP 200 with "status": "error". This is a breaking change for clients that copy datastreams across workspaces. Copies within the same workspace are unchanged, and do not accept the new request fields.
Management API read-only fields¶
With this release, the computed absolute_url and next_run fields of /api/datastreams/{id}/ are read-only. Values sent for them in a PUT or PATCH request are now ignored. These values were never stored, but previously the response echoed them back as though they had been.
Extract API key scopes¶
With this release, API keys need the datastream:read scope to read /api/extracts/ and the datastream:write scope to change extracts.
Self-service Management API keys¶
With this release, on instances where self-service key management is enabled, users who are not administrators can create their own Management API keys, scoped to their permissions. Keys created by non-administrators need an expiry date. Users can always revoke the keys they issued.
Smart Fetch scheduling¶
With this release, the Smart Fetch Earliest fetch start (UTC) time can be set and read through the Management API, on v1 and v2, as the start field alongside the existing deadline. Both scheduling times now reject a non-UTC offset, such as 09:00:00+02:00, instead of storing its clock time as UTC. A UTC time such as 09:00:00Z is still accepted.
Confirm-gated delete actions¶
With this release, the manage_datastream, manage_authorization, and manage_fetch MCP tools have gained confirm-gated delete actions. Before anything is removed, the tool previews the impact, such as dependent datastreams or extracts that will be lost. For more information, see Using Adverity MCP.
Datastream transformation management¶
With this release, the manage_datastream MCP tool has gained actions to attach, detach, and list the transformations on a datastream, so a transformation can be added or removed without risking an existing transformation being silently dropped. For more information, see Using Adverity MCP.
MCP tool reliability¶
With this release, the MCP server’s Management API client now retries a GET request once after a transient network timeout or connection error before surfacing it to the caller, reducing tool failures caused by a single slow backend response.
Shopify¶
With this release, deprecated pre_tax_price fields have been removed from the Shopify connector for the Line items, Line item refunds, and Shipping and fulfillment report types. This cleans up Shopify reports by ensuring that users only see relevant fields.
Resolved issues¶
With this release, issues have been resolved for the following connectors and destinations:
AccuRanker
Hubspot
Klaviyo
Piano
Profound
Shopify
API jobs filtering¶
With this release, a bug has been resolved where the GET /api/jobs/ endpoint returned HTTP 500 for non-numeric stack_id parameters in the Management API. Empty stack_id parameters are now ignored, improving the robustness of the endpoint.
Datastream PATCH requests¶
With this release, a bug has been resolved where PATCH /api/datastreams/{id}/ in the Management API returned HTTP 500 when the request contained app_label or an object-valued stack. Both fields are now read-only, so the request succeeds and these values are ignored. Responses still include stack with the same nested fields as before. To move a datastream to another workspace, keep using stack_id.