Essential Blocks renders pricing tables, accordions, tabs, feature lists, image accordions, and post grids as structured Gutenberg blocks. The output ends up in wp_posts.post_content as block markup that resolves to real HTML at render time. SleekAI reads the rendered page, so a pricing table becomes a list of plans with prices, an accordion becomes a Q&A pair set, and a post grid becomes a small in-page index, all of which land in the chatbot's context.

The data-source wizard lets you also pull ACF fields, custom taxonomies, and any postmeta referenced from a block, so a tabbed product spec sheet built with Essential Blocks tabs can pull from the same custom fields the rest of the theme uses. Display conditions scope different bots to different templates or post types, so a pricing-page bot and a docs-page bot can run side-by-side under multibot without code.

Conversations are logged inside WordPress with model name, token usage, and the page URL, so you can spot where the bot guesses vs. where it answers cleanly, and tighten either the prompt or the underlying block copy. The JS API and PHP embed open the bot from a custom button or a specific block click, with the current page already loaded as context.

Essential Blocks sites are usually iteration-heavy. Marketers ship landing pages, swap pricing toggles, reorder feature lists, rewrite accordion answers, and republish without involving engineering. A chatbot that reads a separate FAQ document breaks that workflow, because every change in the editor needs a mirrored change in the chatbot's training data.

Within a few weeks the bot is answering from outdated copy and producing wrong prices, which is worse than not having a bot at all, especially on pricing and comparison pages where the cost of a wrong answer is high. The structural fix is to ground the bot in the rendered page so the marketing team's source of truth is also the chatbot's source of truth. With Essential Blocks specifically, this is a clean win, because the plugin's blocks render to structured HTML that the bot can parse meaningfully.

Pricing tables become plan data, accordions become Q&A, tabs become labeled sections, and post grids become content indices. Combined with display conditions that scope different bots to different templates, the chat layer matches the structure of the site itself instead of fighting against it. The result is a bot that stays correct without anyone having to maintain it as a separate artifact.