> ## Documentation Index
> Fetch the complete documentation index at: https://nikita-shkoda.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Connect to AI Tools

> Access documentation directly in your IDE and AI tools via MCP

The Model Context Protocol (MCP) enables AI assistants to access this documentation in real-time. By connecting this documentation as an MCP server, you ensure that tools like Cursor, Claude, and Windsurf have access to the latest guides and API references.

<Card title="MCP Server URL" icon="link">
  The MCP server is available at:

  ```
  https://nikita-shkoda.mintlify.app/mcp
  ```
</Card>

## Connection Guide

<Tabs>
  <Tab title="Cursor">
    Open any docs page and use the contextual menu -> "Connect to Cursor".

    Alternatively, you can setup the MCP server connection manually:

    <Steps>
      <Step title="Open Cursor Settings">
        Open Cursor Settings from the top right corner of the editor window, or pressing <kbd>⇧ ⌘ J</kbd> (Mac) or its equivalent on Windows.
      </Step>

      <Step title="Go to Tools & MCP" />

      <Step title="Open MCP servers settings">
        Scroll down to the end of the MCP servers list, and click "+ New MCP Server".
      </Step>

      <Step title="Add server configuration">
        ```json theme={null}
        {
          "mcpServers": {
            "lstpsche docs": {
              "name": "lstpsche Docs",
              "url": "https://nikita-shkoda.mintlify.app/mcp"
            }
          }
        }
        ```
      </Step>

      <Step title="Save the mcp.json file" />
    </Steps>
  </Tab>

  <Tab title="Windsurf">
    <Steps>
      <Step title="Open Cascade tools settings">
        Open the **Cascade** panel, then click the tools plug icon and choose **Manage tools**.
      </Step>

      <Step title="Open raw MCP config">
        In the tools panel, click **View raw config** to edit the underlying JSON configuration.
      </Step>

      <Step title="Add the lstpsche Docs MCP server">
        ```json theme={null}
        {
          "mcpServers": {
            "lstpsche-docs": {
              "type": "sse",
              "url": "https://nikita-shkoda.mintlify.app/mcp"
            }
          }
        }
        ```

        If you already have other `mcpServers` defined, just add the `lstpsche-docs` object alongside them.
      </Step>

      <Step title="Save and reload tools">
        Save the config, then reload tools (or restart Windsurf) so Cascade can connect to the server.
      </Step>
    </Steps>
  </Tab>

  <Tab title="Claude Code">
    <Steps>
      <Step title="Install the Claude CLI">
        Install the official <code>claude</code> CLI from Anthropic (see their docs for the latest installation instructions).
      </Step>

      <Step title="Register the lstpsche Docs MCP server">
        Run this command in your terminal:

        ```bash theme={null}
        claude mcp add --transport http "lstpsche Docs" https://nikita-shkoda.mintlify.app/mcp
        ```

        This tells Claude Code to treat these docs as an HTTP MCP server.
      </Step>

      <Step title="Verify the server is available">
        List configured MCP servers:

        ```bash theme={null}
        claude mcp list
        ```

        and check that <code>lstpsche Docs</code> appears in the output.
      </Step>
    </Steps>
  </Tab>

  <Tab title="Perplexity">
    <Info>
      **Perplexity** does not currently expose a generic MCP client, so it cannot connect directly to the full `/mcp` server.
    </Info>

    <Steps>
      <Step title="Use the contextual menu">
        From any docs page, use the contextual menu and choose <strong>Open in Perplexity</strong> (if enabled in <code>docs.json</code>).
        This opens a Perplexity conversation with the current page content as context.
      </Step>

      <Step title="Ask follow-up questions">
        Ask questions in Perplexity about what you see on the page; it will answer based on that provided context.
      </Step>
    </Steps>
  </Tab>

  <Tab title="ChatGPT">
    <Info>
      **ChatGPT** does not yet provide a general-purpose MCP client, so it cannot connect directly to the full `/mcp` server.
    </Info>

    <Steps>
      <Step title="Open the current page in ChatGPT">
        From any docs page, use the contextual menu and choose <strong>Open in ChatGPT</strong> (if enabled in <code>docs.json</code>).
        This opens a ChatGPT conversation with the page content pre-loaded as context.
      </Step>

      <Step title="Provide extra context when needed">
        For deeper questions, paste additional snippets or links from the docs into the same ChatGPT thread so it can reason over more context.
      </Step>
    </Steps>
  </Tab>
</Tabs>