Selldone

Merchants who run online stores today must log into separate dashboards to update products, build landing pages, manage orders, and handle customer operations, with any automation limited to fixed rules rather than a reasoning agent. Businesses that want a storefront shaped to a specific model, such as a marketplace, booking service, or print-on-demand shop, are often forced to fit a generic commerce template or hire developers to build something custom from scratch. Keeping an AI assistant working across pages, inventory, customers, and finance normally means building custom integrations for each system. The page frames this as needing to 'live in another dashboard' just to keep operations moving.

Selldone Business OS lets a merchant connect ChatGPT, Claude, Codex, or any MCP agent directly to their shop, with access scoped to that shop, revocable at any time, and gated by approval controls for sensitive changes. The connected agent can prepare products, update inventory, build landing pages, and handle approved operations across storefront, orders, and finance, reporting back in the same conversation rather than a separate UI. The page describes three storefront paths built on one shared commerce engine: a ready dynamic storefront with a visual Landing Builder for merchants who want to start selling immediately, an open-source 'Layout' storefront that developers can edit directly and extend with an AI agent through MCP, and Selldone Genesis, which builds fully custom storefront and back-office experiences for models like multi-vendor marketplaces, print on demand, services and booking, or social commerce.

The product is positioned around connecting existing AI agents like ChatGPT, Claude, and Codex to business operations via MCP, treating agent-run operations as a new way to build and manage a store rather than a future feature.