---
lastModified: 2026-11-18
title: Run your stores from your own agent or code.
description: Today you can call Runner tools over MCP and receive store events as webhooks. A REST API, an SDK, and a CLI are planned.
sidebar:
  label: Overview
---

1. **Add the server** — `https://api.runnerai.com/mcp`
2. **Sign in, choose access** — One workspace. The client can always read. Turn on **Make changes** only if the client must change data.
3. **Call list_stores** — Every store tool needs a `store_id` from `list_stores`.

[Start the quickstart](/en/build/quickstart)

## Live today

### [MCP server](/en/build/quickstart)

Use Runner tools from Codex CLI, OpenCode, pi, or an agent that you build. You sign in with OAuth, so you do not need an API key.

`api.runnerai.com/mcp`
  ### [Webhooks](/en/build/webhooks)

  Signed POST requests when orders, carts, customers, reviews, or bookings change.

  `order.placed · cart.abandoned · …`

## Planned

Each tool page will gain these tabs. Its URL stays the same.

### [REST API](/en/build/rest-api) (Planned)

Call the tools over HTTPS
  ### [SDK](/en/build/sdk) (Planned)

  Call the tools from your code
  ### [CLI](/en/build/cli) (Planned)

  Call the tools from a terminal

**Docs for AI agents:** llms.txt · Docs MCP: search_docs, get_page · Copy as Markdown on every guide [Read more](/en/build/docs-for-ai-agents)
