Yes. Product Add-Ons stores per-product add-on groups in product postmeta and global add-on groups as a custom post type. SleekAI maps both via the data-source wizard, so the system message can include the current product's add-on schema and price modifiers alongside the base product fields. The bot only sees the product currently in scope, not the whole add-on library, so the prompt stays short and grounded.

Yes. Each add-on field stores a price adjustment with a type: flat, percent, or per quantity. SleekAI includes the type in the prompt so the bot can compute and explain the right total. If the visitor's question requires arithmetic across a flat base, a percent option, and a per-quantity engraving, the bot can walk through the steps using the same math WooCommerce applies in the cart.

Yes. The required flag on each add-on field is part of the schema SleekAI reads. When a customer asks whether a step can be skipped, the bot answers based on the real rule. That removes one of the most common configurator support questions: customers no longer email asking which fields they truly need to fill before adding to cart.

Yes. File upload add-ons store allowed mime types and max size as field properties. SleekAI exposes those in the prompt so the bot can list acceptable extensions and size limits. The bot does not handle the upload itself; that stays inside the product form so WooCommerce's validation and security checks apply normally.

Yes. Global add-ons are stored as their own custom post type with a category or product assignment. SleekAI reads both the per-product and global add-ons in scope for the current view, so the bot answers correctly whether the engraving option is set on the single product or applied across a 'Personalised gifts' category.

Yes. Multibot mode lets you run a dedicated configurator bot on, for example, the engraving product category with a tighter prompt and tighter presets, while a more general sales bot covers the rest of the catalog. Display conditions scope each bot by taxonomy term, post type, page template, and user role.

No. SleekAI guides, the product form executes. Letting a chatbot push items into the cart with pre-selected add-ons would bypass WooCommerce validation, tax math, and stock checks. Instead, the bot quotes the right configuration and explains how to select it on the existing form, which keeps the audit trail clean and the cart totals trustworthy.

Inside WordPress, per chatbot, with model name, token usage, and page URL on every message. Logs help support spot recurring questions about a specific add-on field, which usually points either to a system-prompt tightening or a label rewrite on the product form itself. Most fixes ship without involving the LLM provider beyond the request itself.