> ## Documentation Index
> Fetch the complete documentation index at: https://platform.atlan.com/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> To act on Atlan objects, use the Atlan MCP server at https://api.atlan.com/mcp or the SDK. To install or publish skills from a terminal, use the `brain` CLI; `brain --map json` prints its command map. The `atlanai` CLI is deprecated. Run a read-only identity check before any write.
> The docs MCP server at /mcp searches these docs only. It cannot read or change Atlan objects.
> SDK packages: Python `atlanai` (PyPI) and TypeScript `@atlanai/sdk` (npm). Show Python first, then TypeScript.
> These docs cover Atlan Agent Registry, Atlan's hosted product. The open-source agent registry specification and its CLI (`brain`) are documented, released and maintained in their own repository; /open-source on this site links to it. Answer general specification or `brain` command questions from the project's own docs, and use these docs only for Atlan-specific setup such as the Atlan adapter, sign-in and workspaces.

# AI Gateway

> Govern how your agents reach tools, with one endpoint, one set of credentials, and one audit trail.

Atlan AI Gateway sits between your agents and the systems they use, so platform and security teams can let agents work without losing control. Agents connect to Atlan once, and the gateway handles sign-in, holds the credentials, checks permissions on every request, and records what happened.

## Gateways compared

AI Gateway is a family of gateways, one for each kind of thing your agents reach:

| Gateway | Agents reach | Status |
| - | - | - |
| **MCP Gateway** | Atlan's platform tools and the remote MCP servers your team registers, such as Linear, Notion, Miro, or Glean | Available |
| **Model Gateway** | Language models, through OpenAI- and Anthropic-compatible APIs | Available |
| **Credential Gateway** | Central management for the credentials agents use to reach other systems | Coming soon |

Every gateway gives your team the same controls:

| Capability | What it means for your team |
| - | - |
| **One endpoint** | Agents connect to Atlan, not to each provider. Adding a provider never touches agent config. |
| **Credentials at the gateway** | Provider secrets are stored encrypted in Atlan and never reach an agent or a client config file. |
| **Workspace access control** | Atlan workspace roles decide who can see and use each provider. |
| **Audit and usage** | Every request is recorded with who made it, why, and how it ended. |

## Core offerings

<CardGroup cols={2}>
  <Card title="Connect client to MCP Gateway" icon="plug" href="/gateway/mcp/how-tos/connect-a-client">
    Add MCP Gateway to Claude Code, Cursor, VS Code, or any MCP client and make your first audited tool call.
  </Card>

  <Card title="Integrations" icon="puzzle-piece" href="/gateway/mcp/how-tos/integrations">
    Put your team's own MCP servers behind the gateway, with their credentials held in Atlan.
  </Card>

  <Card title="Connect client to Model Gateway" icon="microchip" href="/gateway/model/how-tos/connect-a-client">
    Point an OpenAI or Anthropic SDK at Atlan and make your first governed model call.
  </Card>

  <Card title="Track usage and audit" icon="chart-line" href="/gateway/mcp/how-tos/track-usage-and-audit">
    See every tool call across every server: who made it, why, and how it ended.
  </Card>
</CardGroup>

## Choose where to start

* **Want to understand MCP Gateway first?** [MCP Gateway](/gateway/mcp) explains what it adds over connecting each server directly.
* **Looking for a specific tool?** [How tool search works](/gateway/mcp/concepts/find-and-call-tools) shows how agents find the right tool.
* **A call failed?** [MCP Gateway errors](/gateway/mcp/troubleshooting) lists errors by symptom.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.