AI

Changelog

1.5.0 – 2026-10-29

Added

• New Experiment: Content Verification, that adds read-only detection of C2PA Content Credentials in uploaded JPEG/PNG/WebP images (#459).
• New Experiment: Text to Speech, that allows you to generate audio for a specific post and then render audio controls on the frontend (#888).
• New Experiment: MCP Access, that installs and activates the MCP Adapter plugin from WordPress.org when enabled, and adds a Tools → MCP Access screen to control which abilities are exposed to AI agents over MCP (#992).
• New Experiment: WebMCP, that lets an agent browser work in the block editor through document.modelContext, with tools that set the title, insert and edit blocks, save and publish, with every change visible on the page as it happens (#1081).
• New Experiment: Revision Change Summary, that summarizes what changed between two revisions of a post on the visual and classic revisions screens (#1159).
• New Experiment: Block Container Label Generation, that generates concise block container labels in List View and the block Rename modal (#1175, (#1218).
• A post’s Markdown document (?output_format=markdown) has approved comments appended (#1121).
• New background sync processing to handle generating embeddings for posts and terms (#1135).
• “Generate title” button in the post Rename modal (#1140).
• Learn WordPress AI course links to the plugin readmes and the AI settings header for easier access to course resources (#1154).
• A neutral badge icon to thumbnails of images with C2PA data detected in the Media Library grid view (#1164).
• An ai.abilitiesExplorer.fields JavaScript filter for adding read-only table columns to the Abilities Explorer page (#1181).
• Clarified which parts of the plugin are stable Features vs Experiments, flag Experiments that run automatically in Settings > AI, and show when each runs, what it sends to your AI provider, and how many requests it makes from an info icon, the readme, and docs (#1190).
• Featured Connector plugins (Mistral and Vercel AI Gateway) on Settings > Connectors, sorted by name, plus a wpai_featured_connectors filter to change the list (#1212).

Changed

• Update the wording in our capability check notice and change that from an error to a warning (#935).
• Content Translation now translates any block that has rich text instead of just paragraphs and headings (#1009).
• Move the Key Encryption experiment onto the new Secrets API feature plugin with backward-compatible fallback and migration (#1107).
• The /<post>/feed/markdown/ no longer works and serves a 404 (#1121).
• The core/content-query, core/settings-get, and core/users-query abilities now set the public meta flag added in WordPress 7.1 (#1122).
• Summarize enabled and disabled experiments in the AI Status dashboard widget to keep it compact as more experiments are added (#1155).
• Rebuilt the Abilities Explorer on DataViews and added three ai/v1 REST routes to retrieve and run abilities (#1181).
• Replace the “Purge All Logs” button in the AI Request Logs screen with a control to delete logs older than 30 days, 90 days or 1 year, or all logs, and add a before_days argument to DELETE /ai/v1/logs (#1182).
• The core/settings-get ability now matches WordPress core. Settings use their REST API names (for example, title instead of blogname), and a value that does not match its schema is left out. Breaking changes: Some schema fields have changed so ensure you are using the correct field names in your ability calls; the core/read-settings alias was also removed (#1206).
• The core/users-query ability now matches WordPress core. The input is now where, fields, and pagination, and the output is always items and pagination. Users are looked up through the ids, slugs, or usernames lists in where, and lookup by email is no longer supported. Results are ordered by username, a filter that cannot be used is rejected, and a page after the last one returns an error. Breaking changes: Some schema fields have changed so ensure you are using the correct field names in your ability calls; the core/read-users alias was also removed (#1207).
• The core/content-query ability now matches WordPress core. The input is now where, fields, and pagination, and the output is always items and pagination. Posts are looked up through the ids or slugs lists in where, which also requires the post type in type. Posts are filtered by author through where.author, with the author id or slug. Fields are nested, such as title.rendered. A filter that cannot be used is rejected, and a page after the last one returns an error. The author name and slug are only returned to users who may see the author. Only a show_in_abilities value of true exposes a post type. Breaking changes: Some schema fields have changed so ensure you are using the correct field names in your ability calls; the core/read-content alias was also removed (#1208).
• The date and modified fields of core/content-query return site and gmt dates instead of local and gmt. Breaking change: Ensure you are using the correct field names in your ability calls (#1235).
• The registered_date field of core/users-query is replaced by a registered object with site and gmt dates. Note that all 3 core abilities have now identical shape like these shipping in WordPress 7.2. Breaking change: Ensure you are using the correct field names in your ability calls (#1237).

Removed

• The deprecated core/read-settings ability alias. Use core/settings-get with REST API field names (for example, title instead of blogname). The alias could not keep the old field names or response shape after the sync with WordPress core (#1206).
• The deprecated core/read-users ability alias. Use core/users-query with the new where, fields, and pagination input. The alias could not keep the old input and response shape after the sync with WordPress core (#1207).
• The deprecated core/read-content ability alias. Use core/content-query with the new where, fields, and pagination input. The alias could not keep the old input and response shape after the sync with WordPress core (#1208).

Fixed

• AI Request Logging now captures successful generations from providers that use a custom transport, not only those routed through the SDK HTTP transporter (#757).
• Display a clear, context-aware error notice during Ability execution when the configured AI connector is pending approval (#759).
• Remove the text generation capability check from the alt text generation WP-CLI command (#935).
• Feature toggles are loaded with one database query instead of one per feature on sites without a persistent object cache (#1113).
• Content Translation: Translate the blocks inside Post Content when “Show template” is enabled (#1116).
• On multisite, network deactivation restores API keys on every site, and uninstall no longer deletes keys that were still encrypted (#1118).
• The Markdown Feeds and Key Encryption flags are loaded together with the feature toggles instead of with one query each on every request (#1119).
• The site-wide comment markdown feed lists comments instead of the post list (#1121).
• Feed requests are no longer served the post’s Markdown by Accept negotiation or ?output_format=markdown (#1121).
• Accept-header negotiation now follows the header’s quality values and order, so a request that prefers HTML gets HTML and one that prefers Markdown gets Markdown (#1125).
• Ensure canceling meta description generation stops the loading state. Note the actual request won’t be cancelled but the result from that will be discarded (#1128).
• Show an error when a Note fails to save, in place of a “suggestions added” message, and keep reviewing the other blocks when one fails (#1131).
• Skip blocks that are removed, replaced or edited mid-translation (#1137).
• Ensure all strings can be translated in the Abilities Explorer pages (#1146).
• Ensure all strings can be translated in the Content Classification experiment (#1147).
• A comment reply suggestion is no longer inserted into the reply form of a different comment (#1158).
• Match existing terms whose names contain “&” or non-ASCII letters, and add them from suggestions (#1168).
• A non-string term or parent in the AI response no longer fails the whole request when running content classification (#1171).
• Add missing title attribute to singular alternate Markdown discovery link (#1177).
• Show “Regenerate Reply” after a reply suggestion has been generated (#1179).
• Checkbox icon color on the AI settings page (#1183).
• Ensure success and error notices show on the AI Request Logs screen (#1184).
• Key Encryption no longer causes a fatal error when stored API keys can’t be decrypted (#1186).
• Suggest Reply no longer sends the title or excerpt of private or password-protected posts to the AI provider (#1192).
• Comment Moderation no longer automatically analyzes pingbacks, trackbacks or block editor notes (#1192).
• Uninstalling the plugin now removes Comment Moderation analysis data (#1192).
• Omit Author meta line in Markdown feeds when a post has no assigned author (#1210).
• Preserve line-level diff granularity for multi-line blocks like tables and quotes in Revision Change Summary (#1216).
• Markdown Feeds now apply the core the_content_feed and the_excerpt_rss filters, matching RSS and Atom feed output (#1224).
• Ensure the “Serve Markdown via the Accept header” setting can now be saved (#1229).
• The Guidelines prompt no longer contains the HTML entity ' (#1230).
• The permission error for a post now mentions image prompts instead of excerpts when running image prompt generation (#1234).

Security

• Resolved GHSA-7mwf-4ppp-cxmp (GHSA-7mwf-4ppp-cxmp).
• Resolved GHSA-g94j-h78j-rr2j (GHSA-g94j-h78j-rr2j).

1.4.0 – 2026-10-05

Added

• When comment moderation is run, calculate a value score, providing a relevance signal (0–1) for each comment based on how substantive and on-topic it is relative to the post it was left on. Show this score in a new column on the comment screen (#681).
• New experiment, Markdown Feeds, that creates markdown feeds at /feed/markdown/ (available in every feed context) and of individual posts and pages via ?output_format=markdown (#855).
• User-triggered retry for Content Translation if it fails (#941).
• Limit comment moderation bulk analysis queue size (#972).
• Storage and CRUD layer for embedding vectors (new wpai_embeddings table) that can be used to recording the provider and model alongside every vector (#976).
• 9 new default target languages to the Content Translation experiment (#986).
• Guideline categories support to the Content Translation Ability (#987).
• New Vector_Math and Vector_Ranker classes, allowing users to run similarity and ranking queries against vectors (#993).
• Loading indicator for type-ahead suggestion requests (#1029).
• Integrate Alt Text Generation with the Image block’s decorative setting (#1031).

Changed

• Bump minimum supported WordPress version to 7.0.3 (#1001).
• Display specific taxonomy labels in Content Classification notice error messages instead of generic terminology (#916).
• Show a notice and disable title translation when the post title is too short (#967).
• Update to the latest version of the embedding code from the
  PHP AI Client (#975).
• Our generate_embeddings helper function now requires a
  specific embedding model. Pass model (a model ID or a ModelInterface
  instance) and, for a model ID, the provider that offers it. This is a
  breaking change for anyone that happened to start using this function (#975).
• Update Settings vertical ellipsis icon to tool icon (#1001).
• The core/read-content and core/read-users abilities are renamed to core/content-query and core/users-query. The old names keep working as deprecated aliases for now (#1002).
• Use gpt-image-2.5-flare as our default OpenAI image generation model (#1023).
• The core/read-settings ability has been renamed to core/settings-get. The old name keeps working as a deprecated alias for now (#1087).

Deprecated

• The ai/get-post-details ability. Use the single-post mode of core/content-query instead (#1002).

Removed

• Unnecessary URL validation that is now handled by WordPress core (#1011).
• The Enable AI header toggle from the AI settings page (#985).

Fixed

• Ensure focus moves to the Accept button when content generation completes in the Content Resizing modal (#917).
• Increase the default request timeout from 5 seconds to 30 seconds when Core revalidates API keys. If this validation request times out, Core will delete the API key so this is an attempt to fix that (#947).
• Cached the active SEO plugin detection – with a TTL and immediate invalidation on any plugin activation/deactivation – so meta description meta-key lookups no longer re-scan active plugins on every request (#973).
• Improves the Content Translation loading animation to preserve existing text and background colors (#977).
• Settings page failing to render on Gutenberg 23.9+ after @wordpress/dataviews was removed from the private-apis allowlist (#989).
• Ensure our guidelines integration reads published guideline rows from the wp_knowledge post type instead of the removed wp_guideline so those work with the latest version of Gutenberg. We don’t migrate data from the old post type and taxonomy to the new so anyone that had previously set guidelines in an older version of Gutenberg will need to re-add those (#988).
• Import core’s prefixed PSR EventDispatcher in the vendored EmbeddingBuilder (#1005).
• Reserve space for type-ahead text generation when caret is at the end (#1008).
• Updated the settings page container min-height so that it spans the full viewport height (#1013).
• Fix focus style on collapsible card by updating the @wordpress/dataviews package to latest version (#1035).
• AI Request Logs now records HTTP error responses (4xx/5xx) from providers as error with the status code and message, instead of success; Gemini text generation requests are no longer mislabeled as Metadata (#1040).
• Hide the “Suggest Reply” button when the inline comment form is in Quick Edit mode (#1049).
• Editorial Notes only reviewing template blocks when “Show template” mode is enabled in the block editor (#1053).
• Insert the generated summary into the post content when Show template is enabled (#1055).
• Aligned the focus styles of Content Classification suggestion pills with the core Button component (#1060).
• Connector Approvals no longer blocks core’s connector key check by attributing it to the provider plugin or to Gutenberg (#1070).
• Abilities Explorer no longer crashes when an ability’s input property lists several types (#1073).
• Preserve HTML entities and text between shortcodes when normalizing content, while still encoding entity-encoded tags (#1077).
• Prevent options.php validation error when saving general settings with abilities active (#1080).
• Ensure post titles, the site tagline and excerpts no longer contain HTML entities in the markdown feeds (#1086).
• Ensure markdown feeds don’t 404 when the rewrite rules were rebuilt without the feed in place (#1110).

Security

• Ensure we check for the moderate_comments capability before we allow bulk comment moderation and then check for the edit_comment capability for each individual comment that is being moderated (GHSA-pjw7-q94g-4q34).
• Harden the Playground preview publishing workflow to derive pull request and commit metadata from the triggering workflow run (#978).
• Harden JSON data parsing on admin screen (#991).

Older changelog entries can be found in the CHANGELOG.md file.