Rate limits and the daily ceiling now hold under parallel requests.<\/p>","2.3.0":"
Adds an AI notice under every chat, as required by the EU AI Act. It appears\nautomatically after the update; reword it under Appearance \u2192 AI notice if the\ndefault does not suit your site.<\/p>","2.2.1":"
Corrects the suggested privacy policy text, which implied something about your\nprovider's retention that the plugin cannot know. If you copied that text into\nyour privacy policy, re-copy it.<\/p>","2.2.0":"
Adds the [knowscapes_ai_chat] shortcode, placement and access control, starter\nquestions and optional page context. Nothing existing changes: the launcher keeps\nappearing everywhere until you say otherwise, and page context starts off.<\/p>","2.1.0":"
Adds a provider entry for a paid KnowScapes key, and fixes replies stopping\nmid-sentence by raising the reply length limit and reporting when it is reached.<\/p>","2.0.1":"
Fixes a Save button that silently did nothing on the settings screen. Update if\nyour settings would not persist.<\/p>","2.0.0":"
This release removes the shared API key that earlier versions shipped with. After\nupdating, open the settings screen, supply your own key or request a free one,\nconfirm the data transfer and switch the widget back on.<\/p>"},"ratings":[],"assets_icons":{"icon-128x128.png":{"filename":"icon-128x128.png","revision":3758580,"resolution":"128x128","location":"assets","locale":"","width":128,"height":128},"icon-256x256.png":{"filename":"icon-256x256.png","revision":3758580,"resolution":"256x256","location":"assets","locale":"","width":256,"height":256}},"assets_banners":{"banner-1544x500.png":{"filename":"banner-1544x500.png","revision":3758580,"resolution":"1544x500","location":"assets","locale":"","width":1544,"height":500},"banner-772x250.png":{"filename":"banner-772x250.png","revision":3758580,"resolution":"772x250","location":"assets","locale":"","width":772,"height":250}},"assets_blueprints":{},"all_blocks":[],"tagged_versions":["2.3.1"],"block_files":[],"assets_screenshots":{"screenshot-1.png":{"filename":"screenshot-1.png","revision":3758580,"resolution":"1","location":"assets","locale":"","width":1280,"height":960},"screenshot-5.png":{"filename":"screenshot-5.png","revision":3758580,"resolution":"5","location":"assets","locale":"","width":1280,"height":960}},"screenshots":{"1":"The chat panel on the front end, opened from the floating launcher.","2":"Provider selection: your own OpenAI, Anthropic or OpenRouter key, or a free KnowScapes Community key.","3":"Requesting a free key. The transfer is named in full and happens only on this button press.","4":"Placement and access: where the launcher appears, which post types carry it, and whether only signed-in users may chat.","5":"The same assistant placed inline on a page with the shortcode."}},"plugin_section":[],"plugin_tags":[2353,1320,5707,2364,70446],"plugin_category":[41],"plugin_contributors":[287837],"plugin_business_model":[],"class_list":["post-380146","plugin","type-plugin","status-publish","hentry","plugin_tags-ai","plugin_tags-assistant","plugin_tags-chat","plugin_tags-chatbot","plugin_tags-rag","plugin_category-communication","plugin_contributors-xplicator","plugin_committers-xplicator"],"banners":{"banner":"https:\/\/ps.w.org\/knowscapes-ai-client\/assets\/banner-772x250.png?rev=3758580","banner_2x":"https:\/\/ps.w.org\/knowscapes-ai-client\/assets\/banner-1544x500.png?rev=3758580","banner_rtl":false,"banner_2x_rtl":false},"icons":{"svg":false,"icon":"https:\/\/ps.w.org\/knowscapes-ai-client\/assets\/icon-128x128.png?rev=3758580","icon_2x":"https:\/\/ps.w.org\/knowscapes-ai-client\/assets\/icon-256x256.png?rev=3758580","generated":false},"screenshots":[{"src":"https:\/\/ps.w.org\/knowscapes-ai-client\/assets\/screenshot-1.png?rev=3758580","caption":"The chat panel on the front end, opened from the floating launcher."},{"src":"https:\/\/ps.w.org\/knowscapes-ai-client\/assets\/screenshot-5.png?rev=3758580","caption":"The same assistant placed inline on a page with the shortcode."}],"raw_content":"\n
KnowScapes AI Client adds a floating chat assistant to your site. Visitors type a\nquestion, the reply is generated by the AI provider you selected, and the answer\nis rendered as formatted text inside the chat panel.<\/p>\n\n
You choose where the answers come from:<\/p>\n\n
Where the chat appears is up to you: a floating launcher on every page, only on\nthe content types you pick, or nowhere at all \u2014 dropped into a post, a page or a\ntemplate with the Two things make a first message more likely and cost nothing until they are used:\nstarter questions, shown as buttons in the empty chat, and page context, which\nlets the assistant answer about the page the visitor is actually reading. Every\nreply carries a copy button.<\/p>\n\n Every chat shows a short notice that the visitor is talking to an AI and that\nreplies may contain mistakes. You can reword it \u2014 into the language of your site,\nfor instance \u2014 but not switch it off, because the EU AI Act requires that people\ncan tell when they are talking to an AI.<\/p>\n\n Every feature of the plugin is available with every provider. Nothing in the\nplugin is switched off, time-limited or gated behind a payment: which provider you\nuse, and what that provider charges you, is entirely your decision.<\/p>\n\n The Community plan is KnowScapes' free plan. On the KnowScapes side it covers one\nwebsite as the knowledge source plus one additional document, and the free key\nanswers a limited number of messages per hour and per day for the site it is used\non \u2014 at the time of writing 20 per hour and 480 per day, roughly three to six\nvisitor conversations an hour, which suits a small site. A short burst above the\nhourly rate is absorbed rather than refused.<\/p>\n\n When the allowance is used up, the service declines further messages until the\nnext period. Nothing about the plugin changes: no feature is switched off, no\ncountdown starts, and every setting keeps working. Visitors see a short notice\nthat the assistant is unavailable for now, and you see the reason and the current\nallowance on the plugin's settings screen. The current figures \u2014 messages,\npages and document size \u2014 are published at https:\/\/knowscapes.org\/free-key-limits\/\nand the plans at https:\/\/knowscapes.org\/pricing\/<\/p>\n\n You are never obliged to use this key. The same plugin works with your own\nOpenAI, Anthropic or OpenRouter key, or with your own endpoint, and it behaves\nidentically either way.<\/p>\n\n Visitors never see your key. The browser talks only to a REST route on your own\nWordPress site; that route adds the key and forwards the request. The key is\nstored in the WordPress options table and is never printed into a page, a script\nvariable or a response body. The settings screen only ever shows it masked.<\/p>\n\n After activation the plugin is inert. It has no key, no provider endpoint it can\nreach, and no widget on your site, and it never contacts anything on its own \u2014\nnot on activation, not when an admin page loads, not when a visitor views your\nsite.<\/p>\n\n Three things can send a request, and each one is something you start:<\/p>\n\n No visitor data of any kind leaves your site before that last step.<\/p>\n\n This plugin is a client for external AI services. Which service is contacted\ndepends on the provider you select in its settings. Nothing below happens on a\ndefault installation; each transfer requires the configuration steps described\nabove.<\/p>\n\n Used when either KnowScapes Community (free key)<\/strong> or KnowScapes (paid plan)<\/strong>\nis selected. Both reach the same service at the same address; they differ only in\nwhich key is sent, and therefore in the allowance that applies to it.<\/p>\n\n Used only when you fill in the request form on the settings screen, tick the\nconsent box and press the button. It is never called automatically.<\/p>\n\n Used when the provider OpenAI<\/strong> is selected, with your own key.<\/p>\n\n Used when the provider Anthropic<\/strong> is selected, with your own key.<\/p>\n\n Used when the provider OpenRouter<\/strong> is selected, with your own key.<\/p>\n\n When you select the custom provider, requests go to the HTTPS base URL you\nentered and to no one else. You are responsible for the terms that apply there.<\/p>\n\n\n No. KnowScapes issues a free Community key, which is enough to run the widget.\nIf you would rather use OpenAI, Anthropic or OpenRouter, you use your own key\nthere and are billed by them.<\/p><\/dd>\n KnowScapes (paid plan)<\/strong>. It is technically identical to the free entry \u2014 same\nservice, same address, same agents \u2014 but it is a separate entry with a key slot\nof its own, so a paid key is not pasted into a field labelled as the fair-use\ntier, and switching between the two does not overwrite either key. The free-key\nrequest form only appears while the free entry is selected.<\/p><\/dd>\n The KnowScapes service declines further messages until the next period, and\nvisitors see a short notice that the assistant is unavailable for now. The\nplugin itself is unaffected: nothing is disabled, no setting is lost, and the\nwidget resumes on its own once the allowance is back. The reason, and a link to\nthe current figures, appear on the plugin's settings screen while it lasts.<\/p>\n\n If the allowance is too small for your site, enter your own OpenAI, Anthropic or\nOpenRouter key instead \u2014 the plugin works the same way with any of them.<\/p><\/dd>\n The reply ran into Maximum reply length<\/strong> on the settings screen. That setting\nis the room a single answer gets; when a model reaches it, it stops wherever it\nhappens to be, which reads as the assistant breaking mid-thought. Raise it \u2014 the\ndefault is 4096 tokens, roughly 3,000 words, and up to 32768 is accepted. Only the\ntokens actually used are billed, so a generous limit costs nothing by itself.<\/p>\n\n The plugin now notices this rather than leaving you to spot it by reading the\nchat: the reply carries a short neutral line telling the visitor it was cut short\nand that they can ask for the rest, and the settings screen shows which limit was\nhit and suggests a higher one. That notice clears itself once the limit is raised.<\/p><\/dd>\n Use the shortcode Three ways, and they combine. Where the launcher appears<\/strong> can restrict it to\nthe content types you choose, which means single entries of those types \u2014\narchives, the blog index and search results are not single entries and never\nmatch. Never on these entries<\/strong> takes a list of post or page IDs and suppresses\nthe launcher there whatever the rule above says, which is what you want for a\ncheckout or a contact page. And the launcher can be hidden on screens narrower\nthan 600 pixels, for sites where a chat panel would cover the page on a phone. A\nshortcode placed inside an entry always works, exclusions included \u2014 placing it\nthere is itself the decision.<\/p><\/dd>\n With page context on, the title, address and plain text of the page the visitor\nis reading are put into your system prompt wherever you wrote Only fully published, public content is ever read. A draft, a scheduled or\npending post, a private one, a password-protected one, a revision, an autosave\nand anything of a post type that is not publicly queryable are all refused, and\nthe text is taken from the stored content with shortcodes stripped rather than\nexecuted. This matters because the page is identified by an ID that comes from\nthe visitor's browser, so it has to be treated as something anyone can change:\nthe checks are what stop the assistant being asked about post 123 and answering\nout of an unpublished draft.<\/p>\n\n It is off by default, because it changes what leaves your site. Switching it on\nalso adds a paragraph to the privacy policy text the plugin suggests under\nTools \u2192 Privacy<\/strong>.<\/p><\/dd>\n Set Who may use it<\/strong> to signed-in users only. That both hides the chat and\nrefuses the request behind it, so it is not bypassed by someone who knows the\nREST route. Signed-in users also get their own rate limit, counted per account\nrather than per address \u2014 which is fairer as well as more accurate, since\ncolleagues behind one office router no longer share a single budget.<\/p><\/dd>\n An empty box asks the visitor to think of something. Starter questions give them\ntwo to four buttons instead; they cost nothing until one is clicked, and they\nalso steer people towards what your agent answers well. Set them under\nAppearance, one per line.<\/p><\/dd>\n You can reword it, not remove it. The field is AI notice<\/strong> under Appearance;\nleave it empty for the built-in text, or write your own \u2014 in your site's\nlanguage, for example. Article 50 of the EU AI Act requires that people can tell\nthey are talking to an AI, and the same line tells them that replies can be wrong,\nwhich protects you as well. If you build your own front end on the REST route\ninstead of using this widget, show an equivalent notice there.<\/p><\/dd>\n No. The browser only ever talks to a REST route on your own site. The key is\nadded server-side and never appears in any page, script or response.<\/p><\/dd>\n The plugin stores no conversations, no cookies and no visitor identifiers. A\nconversation exists only in the visitor's browser and is gone when the page is\nclosed.<\/p>\n\n That covers this plugin and your site. It says nothing about how long your\nchosen provider keeps the messages you transmit to them \u2014 that is set by their\nterms and privacy policy, it differs between providers, and the plugin has no\nway to know it. Read the policy of the provider you selected before you describe\nretention in your own privacy policy; the links are next to the consent box on\nthe settings screen. The visitor's IP address is hashed into a short-lived counter purely to\nenforce the rate limit. The counter lives in the plugin's own database table,\nexpires after two minutes and is deleted at the next clean-up. The address\nitself is never written anywhere.<\/p><\/dd>\n The request is protected by a WordPress nonce, and nonces have a limited\nlifetime, so a page served from a cache older than that carries a dead one. The\nwidget detects this, fetches a fresh nonce and retries the message once by\nitself, so in practice you should not see this. If you do, reloading the page\nalways fixes it.<\/p><\/dd>\n Visitors are bucketed by the address your web server reports. Forwarded-for\nheaders are deliberately not trusted, because anyone can send one and the limit\nwould be trivial to bypass. Behind a CDN or a reverse proxy, though, that address\nis the proxy, so every visitor shares one budget. If you trust your proxy, supply\nthe real address with the Two limits work together. The per-visitor limit caps how fast one address can\nsend, and the daily ceiling caps how many messages the whole site will answer in\na day, which is the one that matters under abuse because an attacker can change\naddress. Both are on the settings screen; the daily ceiling defaults to 500 and\ncan be set to 0 to remove it. Only messages that actually reach the provider are\ncounted, so malformed requests cannot exhaust your budget.<\/p>\n\n The daily counter is best-effort rather than exact: WordPress offers no atomic\ncounter that works both with and without a persistent object cache, so a burst of\ntruly simultaneous requests can slip a few past the ceiling. It undercounts, never\nover, so it will not lock out your visitors early. Treat it as a guardrail against\nrunaway usage, and set a hard spending limit at your provider as well if the bill\nmatters to you.<\/p><\/dd>\n The chat itself only ever tells visitors that the assistant is unavailable, so\nnothing about your configuration leaks to them. The provider's own wording is\nshown to you when you press Load available entries<\/strong> on the settings screen,\nand, with This version fetches the complete reply and then reveals it progressively in the\nbrowser, which keeps the whole transport on the WordPress HTTP API and needs no\ndirect socket handling. A token-by-token streaming transport is planned for a\nlater release.<\/p><\/dd>\n Two, both GPL-compatible, both shipped unminified, both served locally and never\nfrom a CDN:<\/p>\n\n Nothing in this package is minified and no build step runs over the plugin's own\ncode: every line that runs is a line you can read. Both libraries are the browser\ndistribution each project publishes, unmodified apart from a trailing source-map\ncomment removed because the map files are not shipped. marked is pinned to 15.0.12\nrather than the newest release because from version 16 on it publishes only a\nbundled, minified browser build.<\/p><\/dd>\n\n<\/dl>\n\n\n[knowscapes_ai_chat]<\/code> shortcode instead. Individual entries\ncan be excluded, the launcher can sit in either bottom corner and can be hidden\non phones, and the whole assistant can be kept for signed-in users only.<\/p>\n\nThe KnowScapes Community plan<\/h4>\n\n
Your API key stays on the server<\/h4>\n\n
Nothing happens until you say so<\/h4>\n\n
\n
Features<\/h4>\n\n
\n
External services<\/h3>\n\n
KnowScapes AI API (xplicator GmbH)<\/h4>\n\n
\n
https:\/\/knowscapes.com\/api\/v1\/chat\/completions<\/code> and, when you press the button to load the agent list, https:\/\/knowscapes.com\/api\/v1\/models<\/code>.<\/li>\nKnowScapes free key service (xplicator GmbH)<\/h4>\n\n
\n
https:\/\/knowscapes.com\/api\/trial-keys<\/code> \u2014 the path is the service's own name for its route; the key it issues is a free one with a fair-use allowance, not a time-limited trial.<\/li>\nOpenAI<\/h4>\n\n
\n
https:\/\/api.openai.com\/v1\/chat\/completions<\/code>, https:\/\/api.openai.com\/v1\/models<\/code><\/li>\nAnthropic<\/h4>\n\n
\n
https:\/\/api.anthropic.com\/v1\/messages<\/code>, https:\/\/api.anthropic.com\/v1\/models<\/code><\/li>\nOpenRouter<\/h4>\n\n
\n
https:\/\/openrouter.ai\/api\/v1\/chat\/completions<\/code>, https:\/\/openrouter.ai\/api\/v1\/models<\/code><\/li>\nCustom endpoint<\/h4>\n\n
\n
[knowscapes_ai_chat]<\/code> shortcode and set Where the launcher appears<\/strong> to the shortcode-only option.<\/li>\n<\/ol>\n\n\n\n
Do I need a paid account anywhere?<\/h3><\/dt>\n
I have a paid KnowScapes key. Which provider do I select?<\/h3><\/dt>\n
What happens when the Community plan's allowance is reached?<\/h3><\/dt>\n
A reply stops in the middle of a sentence<\/h3><\/dt>\n
How do I put the chat inside a page instead of in the corner?<\/h3><\/dt>\n
[knowscapes_ai_chat]<\/code> in a post, a page, a block or a\ntemplate. Add a height if the default of 480 pixels does not suit:\n [knowscapes_ai_chat height=\"600\"]. Each shortcode is a chat of its own, so two\non one page do not share a conversation, and the floating launcher \u2014 if you leave\nit switched on \u2014 is a third. To have the chat only where you place it, set Where\nthe launcher appears<\/strong> to the shortcode-only option.<\/p><\/dd>\nCan I keep the chat off certain pages?<\/h3><\/dt>\n
What does page context actually send?<\/h3><\/dt>\n
{title}<\/code>, {url}<\/code>\nor {content}<\/code>, up to the character limit you set, and travel to the provider\nwith each message. {site_name}<\/code> is also available.<\/p>\n\nCan I keep the assistant for my own team?<\/h3><\/dt>\n
Why do so few visitors send a message?<\/h3><\/dt>\n
Can I remove the \"AI assistant\" line under the chat?<\/h3><\/dt>\n
Can visitors see my API key?<\/h3><\/dt>\n
Is any visitor data stored?<\/h3><\/dt>\n
The chat says the page has expired<\/h3><\/dt>\n
The rate limit counts all my visitors as one<\/h3><\/dt>\n
knowscapes_ai_client_client_ip<\/code> filter:<\/p>\n\nadd_filter( 'knowscapes_ai_client_client_ip', function () { return $_SERVER['HTTP_CF_CONNECTING_IP'] ?? $_SERVER['REMOTE_ADDR']; } );\n<\/code><\/pre><\/dd>\nHow do I stop the chat running up a bill?<\/h3><\/dt>\n
Where do I see why the provider rejected something?<\/h3><\/dt>\n
WP_DEBUG<\/code> and WP_DEBUG_LOG<\/code> both enabled, is written to the WordPress\ndebug log. The API key is never included in either.<\/p><\/dd>\nWhy does the reply appear all at once instead of word by word?<\/h3><\/dt>\n
Which third-party libraries are bundled?<\/h3><\/dt>\n
\n
2.3.1<\/h4>\n\n
\n
{prefix}knowscapes_ai_client_rate<\/code>). Before, each counter was read and written back as a transient in two steps, so requests arriving at the same moment could be counted as one and get past a limit. The table is created on activation and on update, and removed on uninstall.<\/li>\n<\/ul>\n\n2.3.0<\/h4>\n\n
\n
2.2.1<\/h4>\n\n
\n
2.2.0<\/h4>\n\n
\n
[knowscapes_ai_chat]<\/code> shortcode, so the chat can sit inside a post, a page, a block or a template instead of only in the corner. Each one is a conversation of its own, and the height is settable per shortcode.<\/li>\n{title}<\/code>, {url}<\/code>, {site_name}<\/code> and {content}<\/code> in the system prompt are filled in from the page the visitor is reading. Off by default. Only fully published public content is ever read \u2014 drafts, scheduled, pending and private posts, password-protected posts, revisions, autosaves and non-public post types are all refused, because the page is identified by an ID that comes from the browser.<\/li>\n2.1.0<\/h4>\n\n
\n
2.0.1<\/h4>\n\n
\n
2.0.0<\/h4>\n\n
\n
1.0.1<\/h4><\/p>\n\n
\n
1.0.0<\/h4>\n\n
\n