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

# Enginy MCP

> Connect Claude Code or another MCP client to Enginy through the hosted remote MCP server.

## Overview

Enginy MCP lets you connect an AI assistant to your Enginy workspace so it can look things up, summarize what it finds, and help you prepare work inside Enginy.

If you are in sales ops, rev ops, or another non-technical role, start with Claude Code first. It is the clearest setup path.

The hosted resource server is:

```text theme={null}
https://openapi.enginy.ai/mcp
```

Behind the scenes, Enginy MCP uses a hosted remote MCP server and a browser sign-in flow. Its tools are generated from Enginy's public API, which helps keep MCP aligned with the rest of the product.

<CardGroup cols={2}>
  <Card title="MCP Overview" icon="plug" href="/mcp/overview">
    Learn how the hosted server, OAuth flow, OpenAPI-derived tools, and workspace policy model fit together.
  </Card>

  <Card title="Connect Claude Code" icon="terminal" href="/mcp/claude-code">
    Register the hosted server in Claude Code, complete OAuth, and verify the granted scopes.
  </Card>
</CardGroup>

<CardGroup cols={2}>
  <Card title="Clients & interfaces" icon="code" href="/mcp/custom-clients">
    Get connection guidance for Claude, Codex, Cursor, VS Code, ChatGPT developer mode, and other remote MCP
    clients.
  </Card>

  <Card title="Client support matrix" icon="table-list" href="/mcp/client-support-matrix">
    Compare supported interfaces, setup shapes, and caveats before choosing the client you want to use.
  </Card>
</CardGroup>

<CardGroup cols={2}>
  <Card title="Supported tools" icon="list-check" href="/mcp/supported-tools">
    See what Enginy MCP can do today, how the tool surface is generated, and what is not guaranteed yet.
  </Card>

  <Card title="Example prompts" icon="message-lines" href="/mcp/example-prompts">
    Use practical prompts for Claude, Codex, Cursor, ChatGPT developer mode, and other MCP clients.
  </Card>
</CardGroup>

<CardGroup cols={2}>
  <Card title="Scopes and policy" icon="shield-check" href="/mcp/scopes">
    Review scope groups, policy ceilings, and how Enginy maps MCP permissions onto OpenAPI routes.
  </Card>

  <Card title="FAQ / known limitations" icon="circle-question" href="/mcp/faq-known-limitations">
    Get quick answers about current limitations and what assumptions are unsafe to make.
  </Card>
</CardGroup>

## When to use MCP instead of direct HTTP

* Use MCP when you want an AI assistant to work inside Enginy for you, without building a custom integration.
* Use direct HTTP when an engineer is building a fixed product integration against specific Enginy API endpoints.

## Fastest path for Claude Code

This is the recommended first setup for most teams.

1. In Enginy, enable MCP for the workspace and allow `http://localhost:3118/callback`.
2. Add the hosted server:

```bash theme={null}
claude mcp add --transport http --callback-port 3118 enginy https://openapi.enginy.ai/mcp
```

3. In Claude Code, open:

```text theme={null}
/mcp
```

4. Authenticate in the browser and approve the requested scopes.
5. Run `mcp_whoami` to confirm the granted client ID, user, and scopes.

<Tip>The full MCP documentation lives in the dedicated MCP tab, not only in this quick-start guide.</Tip>