> **Building with AI coding agents?** Install the authstack plugin with one command. This equips your agent with accurate Scalekit implementation patterns.
>
> **Recommended** (the `-y` flags skip prompts, so the command runs without an interactive terminal):
> ```bash
> npx -y @scalekit-inc/cli setup -y
> ```
>
> Global:
> ```bash
> npm install -g @scalekit-inc/cli
> scalekit setup -y
> ```
>
> Supports Claude Code, Cursor, GitHub Copilot, Codex + skills for other Agent Skills-compatible agents in `~/.agents/skills`.
> Skills: integrate-agentkit, implement-saaskit, add-mcp-oauth, implement-sso, implement-scim.
> [Full setup guide](https://docs.scalekit.com/dev-kit/build-with-ai/)

---

# What is Agent Gateway

Agent Gateway gives your team one MCP URL for Claude, ChatGPT, Cursor and other AI tools, with each person signed in and access set per connection.
**Agent Gateway gives your team's AI tools governed access to the apps your company uses.** You enable it in the Scalekit dashboard and get one MCP endpoint URL. Each person adds that URL to their AI client, such as Claude, ChatGPT, Cursor or VS Code, and signs in with their own account. The tools they see come from the connections you add, filtered by the groups they belong to.

> note: Available on request
>
> Agent Gateway is available on request. [Contact us](/support/contact-us/) to enable it for your
> workspace.

Agent Gateway is for the admins who give their own team access to AI tools, such as IT or security teams. Everything is set up in the dashboard, with no code. Building an agent into your own product for your customers? See [AgentKit](/agentkit/overview/).

## Choose AgentKit or Agent Gateway

Both products use the same [connectors](/agentkit/connectors/). They differ in who builds the agent and who uses it.

|  | AgentKit | Agent Gateway |
| --- | --- | --- |
| You want to | Let **your product's** agent act in **your customers'** apps | Give **your team's** AI tools access to **your company's** apps |
| Who sets it up | Developers, in code | Admins, in the dashboard |
| Who signs in | Your customers, through your app | Your team, with their own accounts |
| The agent | One you build and host | Existing AI clients, such as Claude, ChatGPT, Cursor or VS Code |
| How tools are reached | SDK, REST API or a [Virtual MCP server](/agentkit/mcp/overview/) | One MCP endpoint URL for the whole team |
| Access control | Your code chooses the tools for each call | Groups and per-connection policies |

**Use AgentKit** when you're building a product, such as an inbox assistant or a sales copilot, and your users connect their accounts inside your app. Start with the [AgentKit quickstart](/agentkit/quickstart/).

**Use Agent Gateway** when people on your team already use AI tools and need them to read and act in company apps, with you deciding who can use which tools.

Protecting an MCP server you built yourself? That's [MCP Auth](/authenticate/mcp/quickstart/).

## How it works

1. **You enable the Gateway** in an environment. Scalekit creates its MCP endpoint and adds you as its first user.
2. **You add connections**, such as Gmail, Outlook, Slack or GitHub. These are the same [connections](/agentkit/concepts/) AgentKit uses, set up once for the environment.
3. **Each person connects their AI client** to the MCP endpoint URL and signs in. The client asks for their consent the first time it acts for them.
4. **Each person connects their own app accounts.** The first time a tool needs an app the person hasn't connected, the AI client gives them a link to connect it.
5. **Groups decide which tools each person sees.** On each connection you choose which groups can use it, and whether they get every tool or only some.

Tool calls through the Gateway appear under **Overview** and **Logs** in the dashboard.

## What you set up in the dashboard

| Task                                    | Where                                    | Guide                                                 |
| --------------------------------------- | ---------------------------------------- | ----------------------------------------------------- |
| Enable the Gateway and get its URL      | **Getting Started**                      | [Quickstart](/agentkit/agent-gateway/quickstart/)              |
| Connect Claude, ChatGPT, Cursor and more | The URL from **Settings**                | [Connect AI clients](/agentkit/agent-gateway/connect-ai-clients/) |
| Add users, SSO and SCIM                 | **Identity**                             | [Users and access](/agentkit/agent-gateway/users-and-access/)  |
| Decide who can use which tools          | **Groups**, and each connection's **Policies** tab | [Access policies](/agentkit/agent-gateway/access-policies/) |
| Review plans and limits                 | **Billing**                              | [Plans and limits](/agentkit/agent-gateway/plans-and-limits/)  |

## Next

Enable the Gateway and connect your first AI client in the [Quickstart](/agentkit/agent-gateway/quickstart/).


---

## More Scalekit documentation

| Resource | What it contains | When to use it |
|----------|-----------------|----------------|
| [/llms.txt](/llms.txt) | Structured index with routing hints per product area | Start here — find which documentation set covers your topic before loading full content |
| [/llms-full.txt](/llms-full.txt) | Complete documentation for all Scalekit products in one file | Use when you need exhaustive context across multiple products or when the topic spans several areas |
| [sitemap-0.xml](https://docs.scalekit.com/sitemap-0.xml) | Full URL list of every documentation page | Use to discover specific page URLs you can fetch for targeted, page-level answers |