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

# MCP Server

> Connect your AI coding assistant directly to FrankieOne documentation

## Overview

FrankieOne provides a [Model Context Protocol (MCP)](https://modelcontextprotocol.io/) server that gives AI coding assistants direct access to our documentation, API references, and integration guides.

This means you can ask your AI assistant questions about FrankieOne — like how to structure API payloads, configure webhooks, or implement KYC checks — and get answers sourced directly from our docs, without leaving your IDE.

<Note>
  No authentication is required. The MCP server is a public endpoint available to all FrankieOne customers and developers.
</Note>

## Quick setup

Add the following to your MCP client configuration:

```json theme={null}
{
  "mcpServers": {
    "frankieone": {
      "type": "url",
      "url": "https://docs.frankieone.com/mcp"
    }
  }
}
```

## Setup by tool

<AccordionGroup>
  <Accordion title="Claude Code">
    Add to your `~/.mcp.json` file:

    ```json theme={null}
    {
      "mcpServers": {
        "frankieone": {
          "type": "url",
          "url": "https://docs.frankieone.com/mcp"
        }
      }
    }
    ```

    Restart Claude Code to activate the server.
  </Accordion>

  <Accordion title="Claude Desktop">
    Open **Settings > Developer > Edit Config** and add to your `claude_desktop_config.json`:

    ```json theme={null}
    {
      "mcpServers": {
        "frankieone": {
          "type": "url",
          "url": "https://docs.frankieone.com/mcp"
        }
      }
    }
    ```

    Restart Claude Desktop to activate.
  </Accordion>

  <Accordion title="Cursor">
    Go to **Settings > MCP Servers > Add Server** and configure:

    * **Name:** `frankieone`
    * **Type:** `url`
    * **URL:** `https://docs.frankieone.com/mcp`
  </Accordion>

  <Accordion title="Windsurf">
    Go to **Settings > MCP Servers** and add:

    ```json theme={null}
    {
      "mcpServers": {
        "frankieone": {
          "type": "url",
          "url": "https://docs.frankieone.com/mcp"
        }
      }
    }
    ```
  </Accordion>

  <Accordion title="Any MCP-compatible client">
    The FrankieOne MCP server is available at:

    ```
    https://docs.frankieone.com/mcp
    ```

    Add this as a remote URL-based MCP server in your client's configuration. No API key or authentication is needed.
  </Accordion>
</AccordionGroup>

## Available tools

| Tool                 | Description                                                                                                               |
| -------------------- | ------------------------------------------------------------------------------------------------------------------------- |
| **SearchFrankieone** | Search across the FrankieOne knowledge base to find documentation, code examples, API references, and integration guides. |

### SearchFrankieone parameters

| Parameter          | Type    | Required | Description                                                             |
| ------------------ | ------- | -------- | ----------------------------------------------------------------------- |
| `query`            | string  | Yes      | A query to search the content with                                      |
| `version`          | string  | No       | Filter to a specific API version (e.g., `v2.0.0`)                       |
| `language`         | string  | No       | Filter to a specific language code (e.g., `en`, `zh`). Defaults to `en` |
| `apiReferenceOnly` | boolean | No       | Only return API reference documentation                                 |
| `codeOnly`         | boolean | No       | Only return code snippets                                               |

## Example queries

Once connected, try asking your AI assistant:

* "How do I create an individual entity using the FrankieOne v2 API?"
* "What fields are required for a Brazilian KYC check?"
* "Show me the webhook payload format for check completion"
* "How does forceRefresh work when re-running a verification?"
* "What are the available recipe types for international checks?"
* "How do I configure OneSDK for document capture?"

## How it works

The MCP server connects your AI assistant to FrankieOne's documentation in real time. When you ask a question:

1. Your AI assistant sends the query to the FrankieOne MCP server
2. The server searches across all published documentation
3. Relevant content — including code examples, API references, and guides — is returned
4. Your AI assistant uses this context to provide an accurate, up-to-date answer

<Tip>
  The MCP server always returns content from the latest published documentation, so you'll always get current information about FrankieOne's APIs and features.
</Tip>

## Requirements

* An MCP-compatible AI tool (Claude Code, Claude Desktop, Cursor, Windsurf, or similar)
* Internet connectivity to reach `docs.frankieone.com`
* No API key, authentication, or FrankieOne account required