Spectra layers info boxes, containers, advanced headings, forms, and tabs on top of the WordPress block editor, and stores the result inside wp_posts.post_content using the standard block grammar. SleekAI reads the rendered page output at request time, so info-box headings and descriptions, tab labels and panels, form labels, and any prose inside containers all flow into the chatbot context as structured text.

Display conditions scope each chatbot by post type, taxonomy term, user role, logged-in state, and URL pattern, so the same site can run a pricing-page bot, a documentation bot, and a customer-area bot under multibot. The SleekAI Wizard maps any post meta or ACF fields used by Spectra Pro's dynamic content into named variables the system prompt can reference, even when those values are conditionally hidden on the page.

Every conversation is stored inside WordPress with the model name, token usage, and page URL. The JS API and PHP embed let you trigger the bot from a Spectra button block or a custom layout, and presets plus the guideline filter constrain off-topic chat without extra configuration.

Spectra sites tend to lean on info boxes, tabs, and forms to break content into digestible chunks. A visitor lands on a Spectra landing page, scans the info boxes, opens a tab to compare plans, and then hovers over a contact form trying to decide whether the team will actually reply. A chatbot that does not see any of that has to guess, and the answers it produces drift further from the page as the team iterates on copy.

The drift is invisible to the team because nobody actively audits the bot, and visitors get confidently wrong answers about features that have been renamed or prices that have changed. SleekAI reads the rendered Spectra page so the chatbot's source of truth is the same one the marketing team is editing. Republish in the block editor and the next conversation reflects the new info-box, the new tab panel, or the new form label.

Display conditions let the team scope a different bot per template without losing the design system that Spectra encourages, because the routing is configuration rather than custom code. A logged-in customer bot can read account context from postmeta, a blog bot can lean on the article body, and a sales bot can stick to the landing-page copy. Conversation logs surface the gaps, which feeds back into both prompt tuning and block-level copy edits, so the team improves the page and the bot in the same release.