Glossary

Definitions of key terms used across Runner AI — storefront, sandbox, agent run, and more.

Runner AI
An AI platform that designs, builds, and continuously optimizes a complete e-commerce storefront from a natural-language prompt.
Storefront
The customer-facing online store generated by Runner AI. Hosted on a runnerai.com subdomain (or a custom domain) and powered by a TanStack frontend backed by a Medusa-based backend.
Sandbox
An isolated Daytona-managed development environment in which a generated storefront runs while a merchant is editing or previewing it.
Agent Run
A single end-to-end execution of an AI agent against a project — for example, generating a new section, running an A/B test, or applying an optimization.
A/B Test
An automated experiment where Runner AI compares two storefront variants on real traffic and promotes the winner based on conversion lift.
Project
A merchant workspace that owns a storefront, its history, its active experiments, and the agent runs performed on its behalf.
Workspace
A team-level container that groups projects, members, billing, and shared configuration under a single organization.
Preview
A live, shareable URL of a storefront sandbox before it is published to its production domain.
Backend Pool
A pool of pre-provisioned Railway-hosted Medusa instances. When a merchant creates a store, an instance is dispatched from the pool and bound to the project.
AGENTS.md
A machine-readable site manifest at /AGENTS.md describing how AI agents can install, configure, and use Runner AI surfaces.
Runner AI - Build, optimize, and scale your AI-native store | Product HuntRunner AI - Build, optimize, and scale your AI-native store | Product Hunt

Producto

PreciosRunner AIRunner LabRegistro de CambiosCaracterísticas

Recursos

BlogCentro de ayudaGuía de inicio

Comunidad

ComunidadDiscordReddit

Empresa

Acerca de
© Runner AI
Política de Privacidad
Términos de Servicio
InicioFlyAI CROPreciosBlogSobre Nosotros
Empezar
Empezar