Surfboard Payments Documentation MCP

A Model Context Protocol (MCP) server that provides AI assistants with direct access to Surfboard Payments' documentation.

Overview

This MCP server acts as a bridge between AI assistants and Surfboard Payments' documentation, enabling AI systems to access current, accurate documentation on-demand. This ensures developers receive precise guidance when integrating with Surfboard Payments APIs.

Features

• Documentation Fetching: Retrieve raw documentation files (Markdown, JSON) from Surfboard's servers
• Documentation Index: Access the central documentation index (llms.txt)
• Basic Search: Find relevant documentation paths based on simple keyword queries
• Documentation Caching: Store documentation for 3-4 days, reducing network overhead and improving response times
• Multi-runtime Support: Run using Node.js, Bun

Installation

Local Installation

npm install @surfai/docs-mcp

Global Installation

npm install -g @surfai/docs-mcp

Usage

When Installed Locally

Running with Node.js

npm run start:node

Running with Bun

npm run start:bun

When Installed Globally

# Run directly from anywhere
node-surfboard-documentation-ai

OR

# Run from anywhere
bun-surfboard-documentation-ai

Available Tools

The MCP server provides the following tools:

1. surfboard_fetch_documentation_index

Fetches the Surfboard documentation index.

Parameters:

• bypass_cache (optional): Whether to bypass the cache (default: false)

Example:

{
  "name": "surfboard_fetch_documentation_index",
  "arguments": {
    "bypass_cache": false
  }
}

2. surfboard_fetch_documentation

Fetches a specific Surfboard documentation file.

Parameters:

• path (required): Path to the documentation file
• bypass_cache (optional): Whether to bypass the cache (default: false)

Example:

{
  "name": "surfboard_fetch_documentation",
  "arguments": {
    "path": "/api/orders.json",
    "bypass_cache": false
  }
}

3. surfboard_search_documentation

Searches Surfboard documentation.

Parameters:

• query (required): Search query
• max_results (optional): Maximum number of results to return (default: 10)
• include_snippets (optional): Whether to include snippets in the results (default: true)
• bypass_cache (optional): Whether to bypass the cache (default: false)

Example:

{
  "name": "surfboard_search_documentation",
  "arguments": {
    "query": "process payment",
    "max_results": 5,
    "include_snippets": true,
    "bypass_cache": false
  }
}

4. surfboard_clear_cache

Clears the documentation cache.

Parameters:

• path (optional): Path to clear from cache (if not provided, clears the entire cache)

Example:

{
  "name": "surfboard_clear_cache",
  "arguments": {
    "path": "/api/orders.json"
  }
}

MCP Integration

This MCP server is designed to be used with AI assistants that support the Model Context Protocol. It provides tools for accessing Surfboard Payments documentation directly from the AI assistant.

Configuration

When Installed Locally

When registering this MCP server with your AI assistant platform and using a local installation:

{
  "command": "node",
  "args": ["node_modules/@surfai/docs-mcp/node.index.js"],
  "env": {}
}

Or for Bun:

{
  "command": "bun",
  "args": ["node_modules/@surfai/docs-mcp/bun.index.js"],
  "env": {}
}

When Installed Globally

When registering this MCP server with your AI assistant platform and using a global installation:

{
  "command": "node-surfboard-documentation-ai",
  "args": [],
  "env": {},
  "autoApprove": []
}

Or for Bun:

{
  "command": "bun-surfboard-documentation-ai",
  "args": [],
  "env": {},
  "autoApprove": []
}

If you're having troubles using the above command, you can try running the following command instead:

~ npm root -g
/Users/{{USER}}/.nvm/versions/node/{{VERSION}}/lib/node_modules # The path to the global node_modules directory

Then, you can use the mcp server in the following way:

{
  "command": "node",
  "args": [
    "/Users/{{USER}}/.nvm/versions/node/{{VERSION}}/lib/node_modules/@surfai/docs-mcp/node.index.js"
  ],
  "env": {},
  "disabled": false,
  "autoApprove": []
}

License

MIT