Recourse is an AI chatbot for customer support that answers out of your own content. It reads your published posts, pages and any public custom post types, builds a search index from them, and answers visitors' questions from that index. Every answer carries numbered citations pointing at the pages it came from, so a visitor can check it and you can see where a wrong answer came from.<\/p>\n\n
Nothing is indexed that a visitor could not already read. Drafts, private posts, password-protected posts and post types you did not tick are all left out, and a Retrieval runs on your own server and needs no account and no credential. Generating the answer text does: you supply an OpenAI-compatible endpoint and its key, and you pay whatever that provider charges. Until you do, the plugin makes no external request at all.<\/p>\n\n Under Settings there is a Model Provider picker that fills the endpoint in for\nyou. These all speak the OpenAI chat format and all work:<\/p>\n\n The list is a convenience, not a limit. Anything else speaking the same format\nworks if you type its address in, and the boxes stay editable after you pick.<\/p>\n\n The key belongs in Recourse needs an OpenAI-compatible chat completions endpoint to write an answer. None ships with the plugin and none is contacted until you enter a base URL under Settings, so an unconfigured install makes no external request at all. You supply the base URL and the key, so the service contacted is whichever one you choose.<\/p>\n\n Whichever provider you point it at, that provider's own terms of service and privacy policy govern the data you send them. The settings screen offers the following as one-click choices, and each one is a separate company with its own terms:<\/p>\n\n Ollama is also offered, and is not an external service: it runs on your own server at localhost, so nothing leaves the machine.<\/p>\n\n Any other OpenAI-compatible endpoint works too. Choose \"Other\" and enter its address, in which case the service contacted is the one you named and its own terms apply.<\/p>\n\n On WordPress 6.9 and later the assistant registers itself with the Abilities\nAPI, so other tools on your site can use it, and it can call abilities other\nplugins register once you list them. Nothing is called without you naming it,\nand an ability marked as destructive is refused even then.<\/p>\n\n On WordPress 7.0 and later, if your site already has an AI connector set up\nunder Settings, the assistant will use it and you do not need to enter an\nendpoint or a key here at all.<\/p>\n\n The chat window says \"Automated assistant. Answers can be wrong.\" under the\nassistant's name. The EU AI Act has required that disclosure of visitor-facing\nchatbots since 2 August 2026, and the duty falls on the site rather than on the\nplugin, so the wording is yours to change through the The widget script shipped as Retrieval does, and so does the index. Writing the answer does not: that needs a model endpoint and a key, which you supply.<\/p><\/dd>\n The passages that matched the question, the question, the recent conversation, and the persona text. See the External services section above for the full list. Your site is not uploaded anywhere.<\/p><\/dd>\n Yes. Return false from the The index is built in batches on WP-Cron, so the size of the site is not the limit. The index file is read on each question, so a very large catalogue will want an object cache.<\/p><\/dd>\n Each site in a network has its own settings and its own index. Uninstalling removes the data from every site in the network.<\/p><\/dd>\n\n<\/dl>\n\n\nrecourse_index_post<\/code> filter is there for anything else your site needs to exclude.<\/p>\n\n\n
https:\/\/api.openai.com\/v1<\/code><\/li>\nhttps:\/\/api.anthropic.com\/v1<\/code><\/li>\nhttps:\/\/api.x.ai\/v1<\/code><\/li>\nhttps:\/\/api.deepseek.com\/v1<\/code><\/li>\nhttps:\/\/api.groq.com\/openai\/v1<\/code><\/li>\nhttps:\/\/openrouter.ai\/api\/v1<\/code><\/li>\nhttps:\/\/api.mistral.ai\/v1<\/code><\/li>\nhttps:\/\/api.moonshot.ai\/v1<\/code><\/li>\nhttps:\/\/dashscope.aliyuncs.com\/compatible-mode\/v1<\/code><\/li>\nhttp:\/\/localhost:11434\/v1<\/code><\/li>\n<\/ul>\n\nwp-config.php<\/code> rather than the database if you would rather it did not travel in a backup:<\/p>\n\ndefine( 'RECOURSE_API_KEY', 'your-key' );\n<\/code><\/pre>\n\nExternal services<\/h3>\n\n
\n
\n
Works with what your site already has<\/h3>\n\n
Telling visitors they are talking to software<\/h3>\n\n
recourse_ai_disclosure<\/code>\nfilter. It is there by default rather than absent.<\/p>\n\nSource code<\/h3>\n\n
assets\/recourse.min.js<\/code> is built from the TypeScript sources at https:\/\/github.com\/ibrahimhajjaj\/recourse, under packages\/widget\/src<\/code>. The readable build is shipped alongside it as assets\/recourse.js<\/code>. Build both yourself with npm install && npm run build<\/code> in packages\/widget<\/code>, then take them from dist\/wordpress<\/code>. That is the same source as the general build, compiled without the loader that would fetch a voice runtime from a content delivery network, since a plugin may not fetch code from another host.<\/p>\n\n\n\n
\/wp-content\/plugins\/recourse<\/code>, or install it through the plugins screen.<\/li>\n\n
Does it work without an API key?<\/h3><\/dt>\n
What content leaves my site?<\/h3><\/dt>\n
Can I hide the widget on some pages?<\/h3><\/dt>\n
recourse_show_widget<\/code> filter.<\/p><\/dd>\nHow large a site can it index?<\/h3><\/dt>\n
Does it work on multisite?<\/h3><\/dt>\n
0.3.0<\/h4>\n\n
\n
0.2.0<\/h4>\n\n
\n
0.1.0<\/h4>\n\n
\n