Badges
CODE
1 badge
Plugin Developer
Recent impact
Score weights high-impact work (commits, releases, approved translations, props) at 3x routine activity.
Last 30 days
4contributions
high4
medium0
score12
Last 90 days
15contributions
high15
medium0
score45
Last 12 months
69contributions
high69
medium0
score207
Contributions
Type
June 2026
Jun 30
Tue · 17:00
Plugins
high
Committed [3591782] to Plugins SVN:
This maintenance release fixes a <a class="wiki" href="http://plugins.trac.wordpress.org/wiki/WordPress">WordPress</a>.org packaging problem that could make the previous update fail to install from the <a class="wiki" href="http://plugins.trac.wordpress.org/wiki/WordPress">WordPress</a> dashboard. Updating ensures you receive the latest improvements and security fixes from the recent releases. No settings or data are affected.
This maintenance release fixes a <a class="wiki" href="http://plugins.trac.wordpress.org/wiki/WordPress">WordPress</a>.org packaging problem that could make the previous update fail to install from the <a class="wiki" href="http://plugins.trac.wordpress.org/wiki/WordPress">WordPress</a> dashboard. Updating ensures you receive the latest improvements and security fixes from the recent releases. No settings or data are affected.
Jun 29
Mon · 12:25
Plugins
high
Committed [3590075] to Plugins SVN:
Adds a Test key button to each AI provider API key field so you can confirm a provider key works in one click without spending credits. Adds built-in WPML and Polylang compatibility, so the chat widget's visible text appears automatically in WPML String Translation and can be translated per language without hand-registering each field. Knowledge Role Restrictions now accept a tag by its display name or its slug, with a clearer field label and an actionable error, so entering a real tag by its name no longer returns a confusing "tag does not exist" message. Also fixes Pinecone sites silently ignoring manual Direct Content knowledge entries when one was the best match, so the bot now answers from them as expected. Adds an optional Max Input Length setting to cap how many characters a visitor can send in one message (0 means unlimited, the default, so existing sites are unchanged); when a cap is set, the chat box now shows a language-neutral live character counter that ramps from neutral to amber to red as the visitor nears the limit, with a gentle nudge at the cap, instead of silently refusing further typing. Extends the built-in Web Search option to Google Gemini chat models, which now answer using live Google Search results (previously Web Search only worked for OpenAI models and was a silent no-op for Gemini), using the Gemini API key already configured and at no extra cost. Includes a security hardening fix for the chat intro message setting following a Wordfence advisory. Also improves the Slack live-agent integration: visitors now receive clean clickable links from agents instead of Slack's raw formatting, and the Slack channel shows the visitor's name and email at the top of the conversation. Adds a View indexed content button on each Knowledge entry that opens a read-only panel showing exactly what was indexed for it, including the stored text split into the same chunks the search uses, each chunk's length, and any stored metadata, so you can confirm what a page or manual entry actually saved without changing anything.
Adds a Test key button to each AI provider API key field so you can confirm a provider key works in one click without spending credits. Adds built-in WPML and Polylang compatibility, so the chat widget's visible text appears automatically in WPML String Translation and can be translated per language without hand-registering each field. Knowledge Role Restrictions now accept a tag by its display name or its slug, with a clearer field label and an actionable error, so entering a real tag by its name no longer returns a confusing "tag does not exist" message. Also fixes Pinecone sites silently ignoring manual Direct Content knowledge entries when one was the best match, so the bot now answers from them as expected. Adds an optional Max Input Length setting to cap how many characters a visitor can send in one message (0 means unlimited, the default, so existing sites are unchanged); when a cap is set, the chat box now shows a language-neutral live character counter that ramps from neutral to amber to red as the visitor nears the limit, with a gentle nudge at the cap, instead of silently refusing further typing. Extends the built-in Web Search option to Google Gemini chat models, which now answer using live Google Search results (previously Web Search only worked for OpenAI models and was a silent no-op for Gemini), using the Gemini API key already configured and at no extra cost. Includes a security hardening fix for the chat intro message setting following a Wordfence advisory. Also improves the Slack live-agent integration: visitors now receive clean clickable links from agents instead of Slack's raw formatting, and the Slack channel shows the visitor's name and email at the top of the conversation. Adds a View indexed content button on each Knowledge entry that opens a read-only panel showing exactly what was indexed for it, including the stored text split into the same chunks the search uses, each chunk's length, and any stored metadata, so you can confirm what a page or manual entry actually saved without changing anything.
Jun 19
Fri · 14:25
Plugins
high
Committed [3578784] to Plugins SVN:
New AI Tools (function calling): the chatbot can decide on its own when to use your enabled tools — web search, image generation, uploaded-PDF answers, and add-on actions like WooCommerce — from natural language, with no trigger phrases required (off by default; enable and pick the allowed tools under Actions then AI Tools, where the phrase-matching system is now called Trigger Phrases). Readable, leak-safe AI error messages now cover every provider for both API and connection or timeout failures, so admins get an actionable hint while visitors only see a friendly fallback. Also adds an optional "Start new chat" menu button, auto-clears page caches when you change an AI theme, and fixes multi-bot non-streaming replies using the wrong bot, Gemini-generated images not displaying, and the Slack setup link. Two Claude models Anthropic retired (4 Opus and 4 Sonnet, the May 2025 versions) are removed and any site using one is auto-switched to its current equivalent.
New AI Tools (function calling): the chatbot can decide on its own when to use your enabled tools — web search, image generation, uploaded-PDF answers, and add-on actions like WooCommerce — from natural language, with no trigger phrases required (off by default; enable and pick the allowed tools under Actions then AI Tools, where the phrase-matching system is now called Trigger Phrases). Readable, leak-safe AI error messages now cover every provider for both API and connection or timeout failures, so admins get an actionable hint while visitors only see a friendly fallback. Also adds an optional "Start new chat" menu button, auto-clears page caches when you change an AI theme, and fixes multi-bot non-streaming replies using the wrong bot, Gemini-generated images not displaying, and the Slack setup link. Two Claude models Anthropic retired (4 Opus and 4 Sonnet, the May 2025 versions) are removed and any site using one is auto-switched to its current equivalent.
Jun 11
Thu · 12:06
Plugins
high
Committed [3568800] to Plugins SVN:
Adds Claude Fable 5 — Anthropic's newest flagship model — to the model picker and Content Generator (and fixes Claude Opus 4.7/4.8 as Content Generator models). Adds a Stop button to halt streaming responses mid-answer, an "Images per Article" slider for the Content Generator, and a toggle to hide the Download Transcript menu item. Settings changes now apply immediately on cached sites — no more clearing caches after toggling. Restores in-chat PDF/Word upload, keeps chat-widget text out of your Google search snippets, and rebuilds the translation template so translators no longer lose work on updates. Also includes RTL admin-navigation, satisfaction-prompt, message-loss, and PHP 8.2 fixes.
Adds Claude Fable 5 — Anthropic's newest flagship model — to the model picker and Content Generator (and fixes Claude Opus 4.7/4.8 as Content Generator models). Adds a Stop button to halt streaming responses mid-answer, an "Images per Article" slider for the Content Generator, and a toggle to hide the Download Transcript menu item. Settings changes now apply immediately on cached sites — no more clearing caches after toggling. Restores in-chat PDF/Word upload, keeps chat-widget text out of your Google search snippets, and rebuilds the translation template so translators no longer lose work on updates. Also includes RTL admin-navigation, satisfaction-prompt, message-loss, and PHP 8.2 fixes.
Jun 04
Thu · 12:19
Plugins
high
Committed [3560922] to Plugins SVN:
Rate limits gain a whole-chatbot total message cap (single ceiling across all users and roles, default Unlimited so existing installs are unchanged) with a live "X of Y used" usage readout, a one-click Reset, and a new "Custom…" option for setting any value beyond the 1–100 presets on both the global and per-role limits. Also fixes custom-provider embeddings end-to-end so indexing and queries use the same model (re-index your knowledge base if you had this on), adds an Azure OpenAI quick-start guide, and fixes the feedback "Send" button rendering as an invisible white-on-white block.
Rate limits gain a whole-chatbot total message cap (single ceiling across all users and roles, default Unlimited so existing installs are unchanged) with a live "X of Y used" usage readout, a one-click Reset, and a new "Custom…" option for setting any value beyond the 1–100 presets on both the global and per-role limits. Also fixes custom-provider embeddings end-to-end so indexing and queries use the same model (re-index your knowledge base if you had this on), adds an Azure OpenAI quick-start guide, and fixes the feedback "Send" button rendering as an invisible white-on-white block.
May 2026
May 29
Fri · 13:14
Plugins
high
Committed [3553574] to Plugins SVN:
Adds a guided onboarding wizard for new users plus the latest AI models — Claude Opus 4.8 and Google’s Gemini 3.x lineup — with automatic recovery for models providers have retired (Gemini 3 Pro, Grok 2) or changed (the Claude Opus 4.7/4.8 temperature deprecation), so no site is left on a broken model. Also makes the first chat message reliable behind full-page caches (WP Rocket, LiteSpeed, Cloudflare APO and similar), auto-retries transient provider errors, and makes the chat launcher fully keyboard-accessible.
Adds a guided onboarding wizard for new users plus the latest AI models — Claude Opus 4.8 and Google’s Gemini 3.x lineup — with automatic recovery for models providers have retired (Gemini 3 Pro, Grok 2) or changed (the Claude Opus 4.7/4.8 temperature deprecation), so no site is left on a broken model. Also makes the first chat message reliable behind full-page caches (WP Rocket, LiteSpeed, Cloudflare APO and similar), auto-retries transient provider errors, and makes the chat launcher fully keyboard-accessible.
May 21
Thu · 04:11
Plugins
high
Committed [3540579] to Plugins SVN:
Recommended for any production install behind a full-page cache: eliminates the intermittent "Access denied" error visitors saw on the first chat message under WP Rocket, LiteSpeed, FlyingPress, W3 Total Cache, or Cloudflare APO, and now also auto-tells those cache plugins to never cache the chat AJAX endpoint. Major additions: Custom Provider (OpenAI-compatible) for Ollama / LM Studio / vLLM / llama.cpp / Azure OpenAI; an end-of-session 👍/👎 satisfaction rating prompt with full admin reporting; ACF PDF auto-indexing into the knowledge base; latest-generation models (Claude Opus 4.7, GPT-5.5, GPT Image 2); an Image Quality dropdown for GPT Image; and a bulk DELETE /transcripts REST endpoint. Also fixes a print regression where MxChat's stale print stylesheet was hiding all page content during normal page prints.
Recommended for any production install behind a full-page cache: eliminates the intermittent "Access denied" error visitors saw on the first chat message under WP Rocket, LiteSpeed, FlyingPress, W3 Total Cache, or Cloudflare APO, and now also auto-tells those cache plugins to never cache the chat AJAX endpoint. Major additions: Custom Provider (OpenAI-compatible) for Ollama / LM Studio / vLLM / llama.cpp / Azure OpenAI; an end-of-session 👍/👎 satisfaction rating prompt with full admin reporting; ACF PDF auto-indexing into the knowledge base; latest-generation models (Claude Opus 4.7, GPT-5.5, GPT Image 2); an Image Quality dropdown for GPT Image; and a bulk DELETE /transcripts REST endpoint. Also fixes a print regression where MxChat's stale print stylesheet was hiding all page content during normal page prints.
May 14
Thu · 14:28
Plugins
high
Committed [3532128] to Plugins SVN:
Major release: REST API with bearer-token auth (transcripts read/delete, knowledge push, health), new API Access admin page, redesigned chat header with 3-dot menu and Download Transcript, SEO improvements to Content Generator (rank_math_title and image alt sync), and custom transcript retention (1–3650 days).
Major release: REST API with bearer-token auth (transcripts read/delete, knowledge push, health), new API Access admin page, redesigned chat header with 3-dot menu and Download Transcript, SEO improvements to Content Generator (rank_math_title and image alt sync), and custom transcript retention (1–3650 days).
May 04
Mon · 02:41
Plugins
high
Committed [3521863] to Plugins SVN:
Fixes the WooCommerce Short Description Visual editor and restores scroll-to-top of the user's last message on streaming replies.
Fixes the WooCommerce Short Description Visual editor and restores scroll-to-top of the user's last message on streaming replies.
April 2026
Apr 29
Wed · 15:56
Plugins
high
Committed [3518614] to Plugins SVN:
New: warns you before switching embedding models so you don't accidentally break your knowledge base. Plus richer Landing Pages, a Full Width fix for block themes, and the chatbot no longer shows broken citations for manual knowledge entries.
New: warns you before switching embedding models so you don't accidentally break your knowledge base. Plus richer Landing Pages, a Full Width fix for block themes, and the chatbot no longer shows broken citations for manual knowledge entries.
No contributions in this category yet.
Developer
1 plugin-
MxChat - AI Chatbot & Content Generation for WordPress
2,000+ active installs
