How Reqursor works with your store platform
Why adding support for another store platform does not mean changing the AI. It reads that platform's rulebook instead.
Reqursor Platform's AI has no built-in knowledge of any one store platform. Everything it knows about what a valid store setup looks like for a given platform comes from one place: that platform's connector.
What a connector actually does
A connector is the piece of Reqursor Platform that knows everything specific to one store platform, so nothing else has to. It has two jobs:
- It does the platform-specific work. When a store goes live or is changed, the connector turns the store's setup into the exact steps that platform needs: preparing its settings, installing the right plugins, applying the store's design (colors, fonts, and spacing), and checking the platform's own health signals. See how a change is planned, made, and checked for where this fits into putting a change live.
- It publishes the rules the AI must write within. Before the AI ever drafts a store setup, it reads a published rulebook from the connector: which design settings exist and in what format, which settings and feature switches are valid, what the platform can and can't do, and which page layout blocks it can display.
| What the connector publishes | What it lets the AI do |
|---|---|
| Design rules | Pick brand colors, fonts, and spacing in a form the platform accepts |
| Settings and feature rules | Choose settings and feature switches that actually exist on this platform |
| List of what it can do | Know what this platform can and can't do before proposing anything |
| Page block rules | Build page layouts using only building blocks this platform can display |
Why supporting another platform doesn't mean changing the AI
Because the AI's knowledge of a platform lives entirely in what the connector publishes, not in the AI itself, adding support for a new ecommerce platform doesn't mean retraining or rewriting the AI. It means writing a new connector that publishes its own rulebook. The AI reads the new rulebook and starts producing valid store setups for the new platform, the same way it always has for the platform supported today.
Today
Reqursor Platform supports one store platform today. Other store platforms fit this same design, but none of them are live yet. See Core concepts for the current state of the platform.
Why this matters for your agency
- The AI never invents settings that don't exist. If a value isn't in the connector's published rules, the AI can't propose it; a draft that breaks the rules is rejected before you ever see it.
- Differences between platforms are spelled out, not guessed. If a platform can't do something, it's because the connector's list of what it can do says so, not because the AI got confused.
- The conversation stays the same across platforms. You describe what you want in plain language regardless of which platform a store runs on; the connector makes sure the result is valid for that platform.
Where to go next
How a change is planned, made and checked
Where the connector's platform-specific steps fit into putting a change live.
How the AI writes your store's setup safely
How the AI's draft gets checked against these rules before you ever see it.
Core concepts
Plain-language definitions for store definition, intended setup, and the rest.
Understanding unexpected changes
How Reqursor spots changes nobody asked for in your stores, automatically, on a schedule, and across all your stores.
One store's problem never affects another
Why every store runs in its own separate environment, and how that separation is checked rather than just promised.