Modula stores each gallery as a modula-gallery custom post. The image list lives in the modulaImages postmeta as a serialized array, with each entry carrying an attachment ID, title, description, alt text, link URL, and per-image dimensions. Gallery-level settings (grid type, hover effect, lightbox theme, sorting) live as additional postmeta on the same gallery post. SleekAI maps that structure, so the system message exposes the current gallery's image list and configuration as named variables.

That changes the conversation on any page that embeds a Modula gallery. "How many photos in the wedding gallery?" gets answered from the serialized list. "Does the third image link somewhere?" returns the real URL from the entry, not a guess. "What's the photo of the venue called?" matches on the stored title or description. Lightbox-only galleries get the same treatment as linked grids, because both share the same modulaImages shape.

Display conditions scope the bot to specific gallery templates or post types, multibot keeps a portfolio bot separate from a documentation bot on the same site, and conversation logs capture every photo-related question with model, token usage, and page URL. SleekAI reads Modula; Modula still owns every write, every reorder, and every lightbox setting.

Modula sits on portfolios, lookbooks, case study pages, and editorial galleries across hundreds of thousands of WordPress sites. Visitors scroll, an image stops them, they want to know more. A generic chatbot at the bottom of the page has no way to tell what they were looking at.

SleekAI removes that gap by reading the same data Modula already stores: the modula-gallery custom post, the serialized modulaImages list, the per-image title and description, the link URL, the lightbox configuration. The bot can name an image, quote its caption, point to the right product or project link, and tell a visitor whether downloads are enabled or licensing is by email. That cuts inbox load for photographers and agencies whose previous workflow was a long thread to confirm one image was indeed available for print.

Editors gain a feedback loop, because the conversation log shows which images drove questions and which captions failed to communicate. None of this requires the bot to touch Modula data. SleekAI reads, Modula writes.

Upload, reorder, lightbox theme, and image-size regeneration stay in the plugin where they belong. Combined with multibot, a portfolio site can run a gallery-aware bot on its case study templates and a separate sales bot on the contact page, each with its own system prompt and its own user-supplied key. Add display conditions on logged-in state for client galleries, and the same approach scales from public lookbooks to private proofing pages without leaking access.

The gallery becomes searchable in conversation, not just by scrolling.