Release notes: September 24, 2026

Note: The fixes described in these notes are scheduled to be available to all clients by end of day Thursday, September 24, 2026.

In this release: Taxonomy Usage Report | Bulk Bio Generator | AI Alt Text | Accessibility | Fixes

Master your metadata: introducing the Taxonomy Usage Report

The Taxonomy Usage Report is a centralized workspace designed to help you inspect, audit, and manage how taxonomy terms are applied across site content in real time.

What it is

Immediately available to you (and currently optimized for the Providers content type, with more coming soon), this tool gives you a unified dashboard to review term distribution, clean up improper tags, and manage search relevance directly from Configuration > Mercury Features > Taxonomy Usage Report.

How it works

• Access & Navigate: Head to the Taxonomy Usage Report. Select your content type  on the landing screen, then filter by your desired vocabulary using the top content bar drop-downs.

• Audit Terms (Base Panel): Review a complete breakdown of terms, total tag counts, and site-wide usage percentages featuring color-coded frequency badges and instant live search.

  

• Inspect & Edit (Explore Panel): Select any term in the Base Panel to open three detailed management tabs on the right:

  • Tagged Items: Inspect specific content assigned to a term. Open the side-panel to unassign terms with a single click. (Note: Externally sourced or synced items carry a "Sourced" badge and cannot be modified directly.)

    

  • Co-occurrences: See where terms overlap across different vocabularies (e.g., Cardiology paired with specific conditions). View content sharing both tags—or missing one—and perform instant bulk additions or removals.

    

  • Related Keywords: Open the "Edit keywords for this term" option to add or remove search keywords linked to a specific term.

    

• Rebuild Data: While reports refresh automatically during regular scheduled maintenance, you can trigger instant, real-time updates by clicking Rebuild in the Content Bar, selecting your content type, and confirming Rebuild snapshot.

  

Why it matters

Maintaining clean, accurate taxonomy is essential for site navigation, accurate content delivery, and internal search performance. Instead of manually cross-referencing multiple pages, this feature lets you identify tagging gaps, adjust co-occurring terms in bulk, and optimize keyword connections—saving significant administrative time while keeping your content ecosystem structured and accurate.

Supercharge profile creation: the bulk provider bio generator

The bulk provider Bio Generator is an AI-powered tool designed to streamline the creation, review, and application of provider biographies across your platform in bulk.

What it is

Located in the DXE toolbar at Configuration > AI > AI Bio Generator, this interface gives teams a unified worklist to draft, fine-tune, approve, and publish AI-generated bios for providers missing bio content, while completely safeguarding third-party synced profiles and existing biographies.

How it works

Note: This feature only includes providers that are missing biography text content on their profile. It does not include any providers with content in the biography field, or any providers that have that field configured to be synced from an external data source.

• Access & Filter Worklist: Navigate to the tool to view all eligible providers missing bio text. Use the Status summary for a quick overview, or refine items using keyword/URL path search and status filters (Missing, Pending, Approved).

  

• Generate & Review Content:

  • Row-Level Actions: Click Generate on an individual provider to create bio text, or select Edit to launch a preview modal with live character count validation. Once satisfied, click Approve to mark it ready for publishing.

    

  • Quick Actions (Top): Use Generate all with AI to create draft text for every provider in the table simultaneously, or Apply all approved to publish approved bios site-wide at once.
  • Selection Bar (Bottom): Check specific row boxes to open the floating bulk actions bar, allowing you to Mark as approved, Generate alt text, or Apply to site for selected items only.
• Publish & Remove: Applying a bio pushes the update directly to the provider's profile and automatically clears them from the worklist. (Note: Once applied, subsequent updates must be made on the individual provider node.)
• Automated Guardrails: The system automatically filters out providers with existing bios or those managed via external sync plugins to prevent data overwrites and avoid extra generation costs.
• Engine Improvements: Powered by a unified generation pipeline to ensure consistent prompt and token logic, backed by parallel requests that prevent timeouts and duplicate API calls.

Why it matters

Automate repetitive drafting while retaining complete human oversight to publish fully detailed, high-quality physician profiles at scale—all while safeguarding your synchronized data and eliminating manual effort.

Expanded AI alt text support

The AI Alt Text Generation feature now supports an expanded range of key content types and interactive page components across the platform.

What it is

An automated accessibility enhancement that displays a direct Generate with AI button on supported image fields, enabling instant, accurate alternative text generation right inside your editing workflows.

How it works

• Content Types: Access AI alt text generation directly within image fields across:
  • Health Library Article
  • Events
  • Post
  • Services
  • Standard Layout pages
• Components: Generate alt text seamlessly within interactive components, including:
  • Image
  • Card
  • Video
  • Slideshow
  • Image Gallery
  • Testimonial Carousel
  • Marquee with Featured Image
  • Listing Grid
     

Why it matters

Expanding AI generation to core components and content types helps your team maintain strict compliance, boost SEO, and speed up publishing with a single click.  This reduces the effort of managing complex sites manually ensuring efficient content creation and helping mitigate the risk of accessibility gaps.

Leveling up accessibility: making the digital experience better for everyone

We have released a series of key accessibility enhancements across our components to ensure full compliance with WCAG standards, streamlined screen reader navigation, and a seamless keyboard experience for all users.

Submenu links

• What it is: Updated accessible labels for "View All" links in Provider and Location submenus.
• How it works: Links now dynamically append context (e.g., "View All providers") in screen reader announcements while keeping the visual display clean. Translation and language cache support have also been updated.
• Why it matters: Complies with WCAG 2.5.3 (Speech Input) so voice-control and screen-reader users can accurately target specific navigation links without changing the visual design.

Provider search bar widget

• What it is: Proper heading semantics for autocomplete category titles in the embeddable search bar.
• How it works: Replaced invalid list structures with proper <h2> heading markup for category groups like "Providers," "Conditions," and "Specialties."

• Why it matters: Satisfies WCAG 1.3.1, giving assistive technology users clear section landmarks to jump directly to specific category groups.

  

Accordion components

• What it is: A complete accessibility refactor for the accordion component.
• How it works: Replaced hidden form inputs with native button elements utilizing aria-expanded attributes and native keyboard handling (Tab, Space, Enter).
• Why it matters: Aligns with WCAG 2.1 SC 4.1.2 by correctly announcing accordion names, roles, and open/closed states to screen readers.

Post search active filter count

• What it is: Converted the filter summary text on the Post Search page into a structural heading.
• How it works: Changed summary text (e.g., "2 filters applied") from a plain list item to a heading while retaining exact visual styling.

• Why it matters: Fulfills WCAG 1.3.1 by allowing screen readers to quickly locate active search filters via standard heading navigation.

  

Embedded widget page titles

• What it is: Appended site names to standalone widget document titles.
• How it works: Page titles now format dynamically with built-in characters escaping and automatic cache invalidation upon site renames.
• Why it matters: Meets WCAG 2.4.2 by giving screen reader users immediate context about which site and widget they are interacting with.

Slideshow component

• What it is: Fully rebuilt slideshow component for complete keyboard operability and standardized control logic.
• How it works: Converted controls into native button elements with ARIA region and tabpanel roles, standardized dot/arrow logic, and added a visual fallback guard for hidden controls.
• Why it matters: Ensures screen reader and keyboard users can seamlessly navigate slide content without getting trapped or missing hidden items.

Range sliders

• What it is: Restored high-contrast visible focus indicators for range sliders (e.g., distance filters).
• How it works: Interactive slider controls now display an outline when focused using a keyboard.

• Why it matters: Meets WCAG 2.4.7 by ensuring keyboard-only users can clearly track their position on the page.

  

Services search field layout

• What it is: Adjusted search input layout spacing on the Services page.
• How it works: Optimized padding and spacing for small viewports (320px) and high zoom levels (400%).
• Why it matters: Satisfies WCAG 1.4.10 (Reflow) by preventing placeholder text from overlapping the magnifying glass icon or requiring horizontal scrolling.

Login screen error handling

• What it is: Automated focus management for failed authentication upon login.
• How it works: Upon page reload after a failed login, focus automatically shifts directly to the top-of-form error message block.

• Why it matters: Instantly alerts screen reader and keyboard users to error details without requiring them to hunt through the page layout.

  

Service specialty cards

• What it is: Unique accessible names and corrected heading structures for service cards.
• How it works: Expanded card link labels to include full specialty context while removing redundant phrases like "Click to view," and updated card headings.
• Why it matters: Complies with WCAG 2.4.4 and 2.4.9 by removing generic links so audio announcements are concise, distinct, and easy to navigate.

• Redirect creation for deleted nodes: Resolved an issue where attempting to create a redirect for a deleted node path triggered a 404 error and failed to save. We updated path validation logic to properly handle parameter conversion errors on missing entities, aligning behavior. Content editors can now reliably set up redirects for deleted content paths without encountering system errors.
• Language-specific cooldowns for alt text: Fixed an issue where generating AI alt text for an image in one language mistakenly applied a temporary generation cooldown to all other translations of that same media item. Cooldown timers are now tracked independently per language per image, enabling users to generate English and Spanish alt text without cross-language waiting periods. Cooldown logic for biography and summary generation remains unaffected.
• Mobile filter submit button display: Resolved a layout issue on iOS Chrome devices where the filter submit button was cut off or rendered off-screen. We updated the button's positioning and rendering properties to ensure it remains fully visible and operable, allowing mobile users to apply search filters as expected.
• Map marker clustering & stability: Upgraded the map clustering engine across Location and Provider search maps to eliminate visual glitches, overlapping icons, and map loading failures on page refresh. The update fixes erratic zoom behavior when selecting specific locations and adds a new Provider Search configuration setting that allows site administrators to toggle marker clustering on or off as needed.