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

# MCP Connectivity: Standards-Based Tool Access for Agents

> MCP Connectivity lets Twin and agents call any Model Context Protocol server as a governed capability, exposing custom tools without bespoke integration.

MCP Connectivity brings the Model Context Protocol into SpineWorkspace. Any MCP server (yours or a third party's) becomes an available capability that Twin and domain agents can invoke, with the same authorization, scoping, and receipts as any other capability.

## Why MCP

* **Open standard** for exposing tools to AI agents.
* **Custom integrations** without building a full connector.
* **Reusable across LLMs** and agent runtimes.

## How It Works

<Steps>
  <Step title="Register an MCP server">
    Point SpineWorkspace at your MCP endpoint and authenticate through One Auth.
  </Step>

  <Step title="Capabilities become available">
    Each MCP tool is projected as a SpineWorkspace capability with scope and inputs.
  </Step>

  <Step title="Twin and agents can invoke it">
    Governed calls flow through the same authorization and receipt path.
  </Step>
</Steps>

## Related

* [Capabilities](/platform/capabilities)
* [Connectors](/platform/connectors)
* [Governed Actions](/how-it-works/governed-actions)


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