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

# Using agents

> Connect AI coding agents to the MoonPay Platform documentation using MCP, llms.txt, and contextual code actions.

AI coding agents can search and reference the MoonPay Platform documentation
directly from your development environment. The docs provide four integration
points: an MCP server for real-time search, agent skills for structured
capabilities, `llms.txt` files for bulk context, and contextual code actions on
every code block.

## MCP server

The documentation includes a
[Model Context Protocol](https://modelcontextprotocol.io/) (MCP) server that
gives AI agents direct access to search and retrieve content. Instead of relying
on web search, your agent queries the documentation index for accurate,
up-to-date results.

The server URL is:

```bash MCP server URL theme={null}
https://dev.moonpay.com/mcp
```

### Connect your client

<Tabs>
  <Tab title="Claude Code">
    Run the following command to add the MCP server to Claude Code:

    ```bash Add MCP server theme={null}
    claude mcp add --transport http moonpay-docs https://dev.moonpay.com/mcp
    ```

    By default this adds the server to your local scope. Use `--scope project`
    to share the configuration with your team via `.mcp.json`, or
    `--scope user` to make it available across all your projects.
  </Tab>

  <Tab title="Cursor">
    Add the MCP server to your project's `.cursor/mcp.json` file, or open
    **Settings → Cursor Settings → MCP** and add a new server:

    ```json .cursor/mcp.json theme={null}
    {
      "mcpServers": {
        "moonpay-docs": {
          "url": "https://dev.moonpay.com/mcp"
        }
      }
    }
    ```

    Restart Cursor after adding the server.
  </Tab>

  <Tab title="Claude Desktop">
    Add the following to your Claude Desktop configuration file
    (`claude_desktop_config.json`). On macOS, this file is at
    `~/Library/Application Support/Claude/claude_desktop_config.json`. On
    Windows, it is at `%APPDATA%\Claude\claude_desktop_config.json`.

    ```json claude_desktop_config.json theme={null}
    {
      "mcpServers": {
        "moonpay-docs": {
          "url": "https://dev.moonpay.com/mcp"
        }
      }
    }
    ```

    Restart Claude Desktop after saving the file.
  </Tab>

  <Tab title="Codex">
    Add the following to your Codex configuration file at
    `~/.codex/config.toml`, or to a project-scoped `.codex/config.toml`:

    ```toml config.toml theme={null}
    [mcp_servers.moonpay-docs]
    url = "https://dev.moonpay.com/mcp"
    ```

    You can also add it with the CLI:

    ```bash Add MCP server theme={null}
    codex mcp add moonpay-docs https://dev.moonpay.com/mcp
    ```
  </Tab>

  <Tab title="Other clients">
    Any MCP-compatible client can connect using the server URL above. Refer to
    your client's documentation for setup instructions.
  </Tab>
</Tabs>

## Skills

The documentation publishes a
[`SKILL.md`](https://agentskills.io/) file that describes MoonPay Platform
capabilities in a structured, machine-readable format. Unlike the MCP server,
which responds to individual queries, `SKILL.md` gives an agent a complete
picture of available workflows, required inputs, and constraints up front.

View the generated file at
[`dev.moonpay.com/SKILL.md`](https://dev.moonpay.com/SKILL.md).

### Install skills

Run the following command to add MoonPay Platform skills to your agent's
context:

<CodeGroup>
  ```bash npx theme={null}
  npx skills add https://dev.moonpay.com
  ```

  ```bash bunx theme={null}
  bunx skills add https://dev.moonpay.com
  ```

  ```bash pnpm dlx theme={null}
  pnpm dlx skills add https://dev.moonpay.com
  ```
</CodeGroup>

The skills CLI discovers and installs the `skill.md` file automatically. Once
installed, your agent can reference MoonPay Platform capabilities without
additional configuration.

## Contextual code actions

Code blocks throughout these docs include contextual actions for sending code
directly to an AI tool. Hover over any code block to see options for Cursor,
Claude, and ChatGPT. Each option copies the code along with surrounding
documentation context so the AI tool understands how to use it.

## llms.txt

The documentation provides
[`llms.txt`](https://llmstxt.org/) files for direct content
ingestion by large language models:

| File                                                     | Description                                                    |
| :------------------------------------------------------- | :------------------------------------------------------------- |
| [`llms.txt`](https://dev.moonpay.com/llms.txt)           | A concise summary of the documentation structure and key pages |
| [`llms-full.txt`](https://dev.moonpay.com/llms-full.txt) | The complete documentation content for full-context indexing   |

Use these files to give an AI tool broad context about the MoonPay Platform
without connecting to the MCP server. You can paste the contents into a chat
session or point a tool at the URL directly.