Description
AI House answers customer questions using your own website content, with two answer branches:
- Quick chat — a direct LLM call for greetings and simple questions.
- Deep knowledge — searches a vector database for relevant content, then answers with citations.
On shared hosting the scarce resource is PHP workers, and a chat request that holds one
for several seconds is a real cost. This plugin is built around that limit: every outbound
call has a hard timeout, a circuit breaker, an answer cache, and a quota guard.
Measured cost per chat request: 3 database queries, 1-3 outbound HTTP calls, and zero writes to wp_options.
Providers
- LLM: OpenAI, Google Gemini, OpenRouter (including free models)
- Vector store: Supabase (pgvector), Pinecone, Qdrant
- API keys are stored per provider and encrypted at rest, or supplied as constants in wp-config.php
Knowledge sources
- Any public post type you choose: posts, pages, products, custom types
- File import — Markdown, CSV, JSON and plain text, with preview before importing
- A dedicated Knowledge post type you can edit in wp-admin
- Automatic re-index when content changes
Languages
- The chat widget ships with Vietnamese and English enabled by default. Twelve languages are built in: Vietnamese, English, Simplified Chinese, Japanese, Korean, Thai, Indonesian, French, German, Spanish, Portuguese and Russian.
- Visitors switch language with a button in the widget. The bot answers in the language the visitor actually wrote in — a Vietnamese question typed into a widget sitting on EN comes back in Vietnamese, and the widget follows the visitor’s language. The chosen language is what the answer falls back to when a message carries no signal (“ok”, “iphone”).
- Welcome message, placeholder, system prompt and language instruction are configurable per language.
- The plugin’s own screens are translated, and you choose which translation to use. Settings → General → Interface language picks the language of the AI House screens — the dashboard, the settings form, the notices — independently of the language of your WordPress profile. The choices are read from the translation files that ship with the plugin (English and Vietnamese today), so a locale added to the plugin shows up there without a code change, and “Follow WordPress” is the default.
Admin dashboard
- Overview — messages, sessions, cache hit rate, latency, queue and circuit breaker state
- Conversations — browse sessions and read full message threads
- Knowledge — import from file, manage items, watch indexing progress
- Diagnostics — test connections, inspect the environment and running configuration
Also included
- Floating chat widget and
[mybot]shortcode, light and dark themes - A AI House shortcut on the WordPress admin toolbar, with quick links to the dashboard, knowledge list and settings
- Answer cache keyed per language, with question normalisation
- Visitor feedback (helpful / not helpful) stored per message
- Source citations returned to the widget so visitors can verify answers
- Background queue with automatic retry and backoff
- Automatic daily purge of old conversation logs
- No jQuery dependency on the front end
External services
AI House has no server of its own. It sends nothing to the plugin authors, and it collects no analytics. It does call services that you configure, with your own accounts, and this section lists every one of them, what is sent, and where to read their terms.
The plugin does nothing until you enter a key. With no key configured there is no outbound request of any kind, other than the two self-tests you can press yourself.
AI model provider — you choose one
- OpenRouter — https://openrouter.ai · Terms: https://openrouter.ai/terms · Privacy: https://openrouter.ai/privacy
- OpenAI — https://openai.com · Terms: https://openai.com/policies/terms-of-use · Privacy: https://openai.com/policies/privacy-policy
- Google Gemini — https://ai.google.dev · Terms: https://ai.google.dev/terms · Privacy: https://policies.google.com/privacy
What is sent: the visitor’s question, the recent turns of that conversation, your system prompt for the current language, and — when the knowledge base has a match — the retrieved text chunks. What is not sent: the visitor’s IP address, any WordPress user data, your other plugin settings, or anything from your admin screens.
Embedding provider — only when you index your own content
The same three providers can supply embeddings (OpenAI, Google and OpenRouter). What is sent: the text of the content you chose to index, split into chunks. This happens only when you press Index knowledge, which you do yourself; nothing is indexed in the background without it.
Vector database — optional
- Supabase — https://supabase.com · Terms: https://supabase.com/terms · Privacy: https://supabase.com/privacy
- Pinecone — https://www.pinecone.io · Terms: https://www.pinecone.io/terms/ · Privacy: https://www.pinecone.io/privacy/
- Qdrant — https://qdrant.tech (self-hosted, or the hosted service; no account is needed for the plugin itself)
What is sent: the text chunks and their vectors, plus the question’s vector when a search runs, so the nearest chunks can be returned. If you turn on the optional conversation mirror, the conversation text is copied there too. If you turn on the optional customer request mirror, the requests visitors leave for a human are copied there as well — including their name, email address and phone number. Both mirrors are off unless you tick them.
Webhook — optional, off until you set a URL
If you paste a webhook URL (n8n, Zapier, Make, your own endpoint) and switch on the hand-over form, AI House POSTs the visitor’s question and the contact details they typed — name, email, phone and their note — to that URL. You choose the destination; the plugin has no default and sends nothing when the field is empty.
Your responsibilities
You are the data controller for your visitors’ conversations. Before enabling a provider, read its terms, and tell your visitors what you do with their messages in your own privacy policy. WordPress’s own guidance on this is at https://wordpress.org/documentation/article/privacy/.
Shortcodes
Two shortcodes, for two different kinds of page.
[mybot] — the chat widget, inline. Put it on a support page when you want the conversation itself to be the page.
`
[mybot title=”Support” height=”480″ theme=”light” lang=”en”]
`
Attribute
Default
What it does
title
the bot’s name
The heading above the chat. Set by you, so the script leaves it alone.
height
480
Height in pixels of the inline widget.
theme
light
light or dark.
mode
auto
auto (the plugin decides), quick (always the model) or deep (always your content).
lang
the site’s chat language
Which language the chat opens in, when it is one of the twelve.
[mybot_help] — a help-centre block: search over your own pages, then a way to reach a person. This is the shape of a support front page: a heading, a large search box, results as the visitor types, topic links you choose, and a button that opens the assistant. It searches your published pages and posts only — never the knowledge base you imported, which is private on purpose.
`
[mybot_help title=”How can we help?” subtitle=”Search the guides, or ask the assistant.” placeholder=”Search the help pages” button=”Search” ask=”Ask the assistant instead” topics=”Getting started|/getting-started, Billing|/billing”]
`
Attribute
Default
What it does
title
“How can we help?”
The heading.
subtitle
empty
One line under the heading.
placeholder
“Search the help pages”
Text inside the search field.
button
“Search”
The submit button’s label.
label
“Search the help pages”
What a screen reader announces for the field.
ask
“Ask the assistant instead”
The button that opens the chat.
topics
empty
Comma-separated label and URL pairs, drawn as chips under the box. A refused URL scheme is dropped.
Every visible word is an attribute, so the page can be written in your visitors’ language without touching the plugin. No template is needed: paste the shortcode into an ordinary Page and it renders.
Installation
- Upload the
mybotfolder to/wp-content/plugins/, or install the ZIP through Plugins > Add New > Upload Plugin. - Activate the plugin. Five database tables are created automatically.
- Open AI House in the admin menu, then go to Settings > Get started — it lists the four things still missing, and nothing there is sent anywhere until you add your own API key.
Quick start with no cost
AI House > Settings > Get started walks through the four things a new install needs, and says which are still missing. Nothing on that screen runs a test by itself.
- Step 1 — connect an AI provider. Choose OpenRouter and paste a key from https://openrouter.ai/keys. The key is checked as soon as you paste it. Then press Test connection: the step’s light turns green when the provider actually answers.
- Step 2 — give it something to answer from. Leave Vector store set to “None” to start with: the assistant then answers from the model alone. To answer from your own content, follow Adding your own knowledge below and press Index knowledge.
- Step 3 — prove the knowledge base can be searched. Type a question that is written on your site and press Look this up. A score above your Minimum score is a pass.
- Step 4 — see what it answers from. This step reports whether the knowledge lock is on, and points at the card that shows the file it lives in. It has no switch on purpose — see Answering only from your own content below.
- Visit your site and open the chat widget.
Answering only from your own content (the knowledge lock)
The assistant answers only from your website’s content when a file called the knowledge lock is on your server:
`
wp-content/uploads/mybot/khoa-tri-thuc.md
`
To switch it on, create that file with this block at the top:
`
`
che-do: off on that line pauses the lock while keeping the file; **deleting the file** removes the lock completely. That is the whole interface — there is no button, no admin code, and nothing in this plugin ever writes to that file, so a plugin update cannot change it.
With the lock on, everything is searched against your knowledge base, and a question with nothing above your Minimum score is refused without calling the model, so it costs no tokens and the visitor is offered the “leave your email” form instead. Short greetings are still answered. A file with no MYBOT-LOCK block is not a lock file; a file that declares a lock but cannot be read properly (a missing or unknown che-do) is treated as on, with the reason shown on the card.
⚠️ A new install does not arrive locked (changed in 1.10.0: earlier versions created this file for you and handed over an admin code; that code was removed at the request of the plugin’s author after it locked people out of their own setting). Until you create the file, the assistant still answers questions your content cannot cover, from the model’s own general knowledge, on your API key.
Adding your own knowledge
- Create a Supabase project in the region closest to your hosting
- Run the SQL file from the project
sql/supabase-schema.sqlin the SQL Editor - Go to AI House > Settings > Knowledge base, choose Supabase, and paste your project URL and service_role key
- Go to AI House > Knowledge, drop in a file, click Preview, then Import
- Go to Settings > Get started and press Index knowledge, then let WP-Cron process the queue
What works without an account of ours
You do not need an account with us, and there is no paid tier of this plugin. AI House has no server of its own: it is free, complete, and the code you install is the whole product. No feature is locked, no trial expires, and installing it sends nothing to the authors.
What it does need is your own account with the provider you choose — an API key for the AI model, and optionally a vector database. That is your account, your key, and your bill, and the plugin talks to those services directly from your site.
FAQ
Does this train or fine-tune a model?
No. It uses RAG. Your content is split into chunks, converted to vectors, and stored. When
a visitor asks something, the closest chunks are retrieved and placed in the prompt.
Nothing is trained, updates are instant, and you can see which source produced each answer.
Can I use it without paying for AI?
Yes. OpenRouter offers free models and Google Gemini has a free embedding tier. Note that
OpenRouter free models are limited to 20 requests per minute and 50 per day until you have
purchased 10 credits. AI House checks your remaining quota before calling and returns a
friendly message instead of a raw error.
What happens if the AI provider goes down?
The circuit breaker opens after a configurable number of consecutive failures and serves
your fallback message immediately, without holding a PHP worker. The real reason is
recorded and shown on the dashboard.
Where are conversations stored?
In the {prefix}mybot_messages table, written in batches by a background job rather than
during the visitor request. Recent messages are sent as context for follow-up questions,
and old logs are purged automatically.
Does it work on multisite?
Yes. Each site gets its own tables, options and knowledge base.
Will my knowledge pages appear in Google?
No. The knowledge post type is not public, so it creates no front-end pages and is
excluded from search. Citations therefore show the title only. If you want linked
citations, index regular pages or posts instead.
Which file formats can I import?
Markdown, CSV, JSON and plain text. PDF and DOCX are not supported because extracting
text from them requires a PHP library that most shared hosts cannot install. Convert those
documents to Markdown first.
Screenshots





Changelog
1.22.12 (13/11/2026)
- The chat box had no minimum height, so before the first message arrived — and on a short phone screen — it collapsed to a narrow strip.
.mybot-widgetnow carriesmin-height: min(520px, calc(100dvh - 110px)): a floor of 520px that is itself capped by the screen, because a floor taller than the space left would push the input box off the display, which is worse than no floor at all. - The shortcode
[mybot]keepsmin-height: auto, so an embedded chat is not turned into a 520px grey wall in the middle of an article. - No PHP change, no new string, no database change — install over the top. This package also contains 1.22.9, 1.22.10 and 1.22.11.
1.22.11 (13/11/2026)
- On a phone the chat panel now fills the screen between the notch and the home bar. The panel used to be anchored on one edge, so its height came from a number in the stylesheet; on a small screen the bottom edge sat under the address bar. At 480px wide and below it is anchored on both edges, the message list takes the space that is left, and every edge adds
env(safe-area-inset-…)so the home bar and the notch cannot cover a button or the last line. dvhin place ofvh, with thevhline kept in front of it as a fallback for browsers that do not knowdvh.- The floating button can no longer be squashed into an oval. It is a flex item, so it could shrink on the cross axis, and a circle drawn on a box that is no longer square is an ellipse. It now has
aspect-ratio: 1 / 1,min-width/min-heightandflex: 0 0 auto, which makes its shape independent of the box around it. - ⚠️ What CSS cannot fix, said plainly: on iOS the layout viewport does not shrink when the on-screen keyboard opens, so a
position: fixedpanel stays where it is and the keyboard covers the input box. That needsvisualViewportin script, orinteractive-widget=resizes-contentin the theme’s viewport tag, and it is not in this release. - No PHP change, no new string, no database change — install over the top.
1.22.10 (13/11/2026)
- The assistant may no longer promise a form. It was told to say that the form under its reply takes the visitor’s details, and it did — even when no form was drawn, because the widget only draws that form when the site decides a person is needed. It is now told not to describe, promise or point at anything “below” or “under” its answer, and that the details area appears by itself when the site decides it should.
- “Can I meet the support team” now opens the hand-over form. Phrasing added: meet the support team · meet the team · meet your team · meet a member · talk to the team · contact the team · gặp nhân viên · nói chuyện với nhân viên · liên hệ nhân viên. There is deliberately no bare “support team”: “does the support team handle warranty claims” is a policy question and must not get a form under it.
- No new string, no setting, no database change — install over the top.
1.22.9 (13/11/2026)
- A new filter lets an add-on answer, not only edit the prompt:
mybot_chat_reply. It runs immediately after the model call and may return the whole reply, so an add-on that knows the answer — a warranty expiry date, a stock level, an order status — answers deterministically: no model call, no tokens, no drift. The reason it exists: the core’s own instruction “never ask the visitor to type an email address or a phone number into the chat” kept winning over facts an add-on had already put in the prompt, four add-on releases in a row. - A return value that is not an array is ignored, so a half-written add-on cannot blank the chat box of a whole site.
- With no filter attached the reply is byte-for-byte what 1.22.8 produced. No new table, no new option, no database change — install over the top.
1.22.8 (04/11/2026)
- The customer requests can now be copied to Supabase, with their own switch. Requests have always lived only in WordPress, on the Requests screen, and that was deliberate: they carry the customer’s name, email address and phone number. If you want them in Supabase as well — for a support inbox that reads from there, or for reporting — tick Also copy customer requests to Supabase on the Knowledge base tab. It is off until you do, and it is a separate switch from the conversation mirror because the two carry different data.
- Run the
mybot_requestsblock insql/supabase-schema.sqlfirst. It creates the table withwp_idas the key — the WordPress id, so a retried copy cannot produce duplicates — plus an index on the date and on the status, and row level security switched on with no policy, which means only theservice_rolekey can read or write it. - Marking a request answered or closed now follows it to Supabase, so the copy does not drift from the record you work from.
- It fails like the conversation mirror, and independently. Three failures in a row switch the request mirror off — and only that one; a broken request table cannot stop conversations from being copied, and the other way round. The reason Supabase gave is printed on the Knowledge base tab and on the Status card, and Test connection writes one probe row into
public.mybot_requestsand removes it again, which is the fastest way to see a missing table or a refused key. - No change to what WordPress stores: every request stays in your own database whatever this switch says.
1.22.7 (04/11/2026)
- The screen that lists the questions visitors left now opens with the conversations they came from. A “Recent conversations” box shows the last few conversations with their message count, their last activity and how many unanswered requests belong to each, and the requests are listed underneath on the same screen. Until now conversations lived only in the dashboard and requests only on this screen, so nothing connected the person who asked to the conversation they were in.
- A new request announces itself in the admin menu. The AI House menu carries a count bubble — the same shape WordPress uses for comments and plugin updates — counting the requests nobody has answered yet. It is visible on every admin screen, it is read aloud as part of the link, and it disappears by itself when you mark a request answered or closed, so it cannot disagree with the list. The Requests screen also says how many are waiting, with a link that filters to them.
- The assistant now offers to take the visitor’s details instead of waiting to be asked. That offer has existed since 1.17.0 but waited for three visitor messages; it is now one.
add_filter( 'mybot_contact_invite_after', fn() => 3 );restores the old patience and0switches the offer off. - And the offer finally has a sentence. Its wording was read from a string key that was never in the language table, so since 1.17.0 the bubble inviting a visitor to leave their details has been drawn with no words at all — two buttons and nothing above them. The sentence now exists in all twelve languages.
- No setting was added and there is no database change.
1.22.6 (04/11/2026)
- The Sources list under an answer printed the same document several times. A long page is stored as several chunks, and a question usually matches several chunks of the same page, so one page appeared four times — which reads as four documents and makes a correct answer look padded. The list is now one row per document: duplicates are merged using the post id the indexer stores, then the link, then the title, and the first row is kept because the chunks arrive in score order and the first is the best evidence.
- Nothing is hidden to tidy the list. A different page still gets its own row, two pages that merely share a title stay separate rows, and a source that carries no id, no link and no title is shown rather than merged away. The chunks are still sent to the model as context — they were never four citations, only four pieces of one document.
- No setting, no new string, no database change, and no need to re-index: this is the step that builds the list when an answer is returned.
1.22.5 (04/11/2026)
- The floating chat button now draws attention to itself. The logo on the button gives a short shake — a small rotation and swell, about a second long — and then waits eight seconds before doing it again, so the button is still for most of every cycle instead of moving all the time. A button that never stops moving makes a page feel broken, and it is the first thing a visitor with vestibular sensitivity meets.
- Three things stop it, and all three matter. If the visitor’s operating system asks for less motion, the button does not move at all. Pointing at or focusing the button pauses it, because aiming at something that is moving is worse than a button that is still. And it does not move while the chat is open, where the logo has become a close cross — a shaking close button says “act now” about the wrong thing.
- It stops for good once the chat has been opened. The animation class is removed rather than paused, so a visitor who has already found the chat is not shaken towards it on every later page.
- Sites that would rather not have it can turn it off in one line:
add_filter( 'mybot_launcher_vibrate', '__return_false' );. It is a filter rather than a setting because a moving element on every page is a matter of taste, not a switch to hunt for. - The animation is applied to the logo, not to the button, so the button still lifts under the pointer on hover. No setting was added, no new string, and no database change.
1.22.4 (04/11/2026)
- A correct add-on licence key was reported as “not a key”. The Add-ons card stopped accepting a key at 128 characters, and a signed key is about 428 — its signature alone is 342. The shape rule also allowed no dots, and every signed key contains two, so correcting the length alone would have left the same refusal worded differently. The limit is now 1024 characters, and letters, digits, dots, dashes and underscores are accepted.
- The card can now give a real verdict — and still verifies nothing itself. It asks the add-on that issued the key, through the new filter
mybot_addon_key_verdict, and prints that answer with the add-on’s name: the signature, the domain and the expiry are all read from inside the key, on your own site, with no request leaving it. With no add-on installed that checks keys, the card says so in words, exactly as before. - A newly saved key clears the previous verdict, so pasting the right key after the wrong one no longer leaves the refusal on screen for five minutes.
- No setting was added or renamed and there is no database change: the verdict is still stored as codes rather than sentences, so the screen follows the language you are reading it in.
1.22.3 (28/10/2026)
- A visitor who asks for a person now always gets a form, even with the hand-over feature switched off. The card form needs no setting, so the widget opens it by itself in that case; with the hand-over feature on, the question form still appears under the answer.
- The instruction now matches that behaviour: when there is no hand-over form, the bot tells the visitor to press the contact icon at the left of the message box instead of saying it cannot help.
- The contact icon is larger — 24 px instead of 20 px in a 40 px button, which is what the owner asked for after seeing it on a phone.
- No setting, no new string, no database change.
1.22.2 (28/10/2026)
- Asking for a person now always offers the hand-over form. Until now that form appeared only when the bot FAILED — the model said it did not know, or your content matched nothing. A visitor who typed “I want to talk to a human” usually got a perfectly good answer from your knowledge base, so nothing was offered, even though the reply often ended with “you can leave your email”. The request for a person is now detected in all twelve languages, on the visitor’s own message, and it wins regardless of how good the answer was.
- The bot is also told what it may promise. With the hand-over form switched on, it says a form below the reply takes the details; with it switched off, it says plainly that it cannot pass the request on and points at your contact page. In both cases it is forbidden to ask the visitor to type an email address or a phone number into the chat — which is how an address ends up stored as a message.
- Matching ignores accents, so “tôi muốn gặp người hỗ trợ” and “toi muon gap nhan vien” are the same request, while ordinary warranty or support questions are deliberately not matched. The filter
mybot_human_intent_phraseslets a site add its own phrasing.
1.22.1 (28/10/2026)
- The contact button in the message box is drawn as a vector glyph, not an emoji. It used to print the person emoji, which each visitor’s operating system draws its own way — a blue person on Windows, a grey silhouette on Android, another weight on macOS — so one button looked like three different buttons. It is now two overlapping chat bubbles, the shape of the plugin’s own mark, crisp at any size and taking the site’s primary colour.
- Its accessible name is unchanged: the string from the language table, in all twelve languages. The glyph itself is hidden from screen readers, so nobody hears it read aloud.
- No behaviour change, no option, no database change: the same button opens the same form.
1.22.0 (26/10/2026)
- Cited links for knowledge imported from a file. Citations have always carried a URL and the widget has always drawn it as a link — but the URL came from
get_permalink(), which is empty for the private knowledge type and for every item imported from a file, so a citation showed a title with nothing to click. Each knowledge item now has its own link, set in a Citation link box on that item’s edit screen or by aurlcolumn in the import file (all three templates include it). - A link is accepted only inside this site: a path such as
/dang-ky, or a URL on the same host includingwww.. Another domain,javascript:, and an empty value are refused, and the check runs when the link is read as well as when it is saved, so a bad URL already in the database cannot reach the widget. On a customer’s site, “the same host” is that customer’s own site. - No widget change, no database change: the stored link travels through chunk metadata into the sources list the widget already renders. The knowledge list also shows a Link column, so an item without one is visible without opening it.
1.21.0 (18/10/2026)
- This plugin’s name changed. The first submission was rejected because “MyBot” was too close to an existing chatbot product called MyBot.chat. It is now AI House Knowledge Chat, the slug is
aihouse, and the text domain, plugin header and translation file names all follow it. If an older version is installed, delete that folder after activating this one. - The description no longer compares this plugin to others, and a dead link to a Qdrant terms page was removed. Twenty translator comments were added and five exception messages annotated, both at the request of the review team.
- No setting was renamed and nothing else changed: options, knowledge base and conversations are untouched.
1.20.0 (18/10/2026)
- New: a help-centre shortcode,
[mybot_help]. A support front page needs three things in one block: a heading, a search box over your own pages, and a way to reach a person. That is now one shortcode with seven attributes, and every visible word of it is yours to write in your visitors’ language.[mybot]still embeds the chat itself, inline, at whatever height you give it. - The search finds your published pages and posts, and nothing else. It never returns the knowledge base you imported — that table is private by design, and a public search box is exactly where an internal price list or a policy draft would leak. Draft, private and password-protected pages are excluded by the query itself, not by a list that could drift.
- Nothing on the help page can be turned into markup. The title, the labels and the topic links are escaped when printed, a
javascript:topic link is refused, results are written withtextContent, and the excerpt the server sends has its tags and shortcodes stripped — so a page title containing a script is shown as text. - A search box that cannot be used as free workload: at least two characters, at most 120, six results, and the same rate limiter as the chat. One character returns nothing and says why.
- No new option, no database change: install over 1.19.0, and a page that already uses
[mybot]keeps working unchanged.
1.19.0 (18/10/2026)
- The directory gave this plugin the slug
aihouse, notmybot, and WordPress.org requires the text domain to equal the slug. The first submission scan came back with about three hundredTextDomainMismatcherrors; they were all one fact. The text domain is nowaihousein 976 places: 656 PHP strings, 160 in the admin sources, 160 in the built bundle, the plugin header,load_plugin_textdomain(), theplugin_localefilter, the translation file names, and the Jed key the dashboard reads. - Renamed by tokenizer, not by search-and-replace. 674 bare
'mybot'literals exist in this codebase and most of them have to stay: the[mybot]shortcode, the admin page slug, the cache group, the menu parent, the Supabase/Pinecone namespace, the folder the knowledge file lives in,function_exists( 'mybot' ). Only the ones that are the text-domain argument of a gettext call were touched — 18 were deliberately left alone, and the tool lists them so a person can check. Plugin URIno longer points atexample.com. The automated scan rejects that domain outright; it now points at this plugin’s own page on WordPress.org. Replace it with your own homepage when you have one — a unique https URL is what the directory asks for.- Twenty
translators:comments added and five exception messages annotated. The scanner treats a placeholder string without a translator comment as an error, and an exception built from a variable as unescaped output. Both were fixed at the source: the comments now say what each placeholder means, and eachthrowcarries a one-line reason explaining that an exception message is not HTML. - New guard:
tools/check-textdomain.mjs— nine rules and four injected-fault controls. It exists because nineteen guards all asked “do these parts agree with each other”, which they did, and none asked “do they agree with the slug” — a value that was consistent everywhere and wrong everywhere. - No new option, no database change: install over 1.18.4. Your saved settings are untouched — option names did not change, only the text domain used for translations.
Older releases
WordPress.org keeps at most 5,000 words of changelog and drops the rest, so the releases before 1.19.0 are one line each here. The full notes for every one of them are in CHANGELOG.md in the repository.
- 1.18.4 — five screenshots taken on a running site, and the
== Screenshots ==section back. - 1.18.3 —
Contributors:names the real account (vunguyenbh) instead of a placeholder. - 1.18.2 — the submission copy checked against the directory’s own rules: five tags, a short description under 150 characters, no Screenshots tab before the pictures existed.
- 1.18.1 — the security review before submission, two hardening fixes, and a check that encodes what a reviewer looks for.
- 1.18.0 — fixed a refusal: a thank-you was answered with “outside the information this website holds”.
- 1.17.0 — the chat can collect a visitor’s contact details and you can read them on a screen of their own. One new table, applied on upgrade.
- 1.16.0 — all the tools moved to the left of the message box and Send became a round arrow.
- 1.15.0 — the message box redrawn: flat icons in a row, the field a filled pill, every control the same height.
- 1.14.1 — fixed: in 1.14.0 the attach-a-picture button opened the file picker, and choosing a photo then did nothing.
- 1.14.0 — the attach-a-picture button: a visitor can send a photo with the question, stored privately on your own site. One new column.
- 1.13.0 — a microphone in the chat — the visitor speaks, the words arrive in the box to check before sending. No recording is stored.
- 1.12.1 — fixed (important): 1.12.0 could take a whole site down when the plugin folder was only partly updated.
- 1.12.0 —
Test connectionreports whether the chosen model can read an image, and message content can carry one. - 1.11.1 —
Knowledge basemoved up to sit directly underGet started. - 1.11.0 — Settings became a sidebar of nine sections opening on
Get started, plus an add-on key field. - 1.10.0 — the knowledge lock lost its switch and its admin code: the file alone turns it on and
che-do: offpauses it. - 1.9.10 — fixed: switching the knowledge lock off with a code that did not match printed a PHP warning.
- 1.9.9 — fixed (critical): 1.9.8 broke every wp-admin screen.
- 1.9.8 — the add-on framework: a registry, a licence layer, and the Add-ons screen.
- 1.9.7 — a
Get startedtab with four steps, each with a status light. - 1.9.6 — a
Connectionstab in Settings. - 1.9.5 – 1.9.0 — the knowledge lock (answer only from your own content), its admin code, the button text and the invitation bubble, and every widget string following the visitor’s language.
- 1.8.0 – 1.6.0 — the admin dashboard became translatable, answers began following the language the visitor actually wrote in, the question box grew with the question,
Test connectionstarted testing the embedding key, the Requests screen and the plugin logo arrived, and indexing learned to read its own vectors back. - 1.5.9 – 1.3.0 — knowledge import from Markdown, CSV, JSON and text; the model lists and the OpenRouter embedding path; the lookup test and the similarity threshold; the Supabase conversation mirror; and the hand-over form that lets a visitor leave their name and phone number.
- 1.2.0 – 1.0.0 — the first releases: the settings screens, the provider and model selection, and the chat widget itself.
