AI Assistant Integration

ByteMCP works with any MCP-compatible AI assistant. This guide covers integration options.

Supported Clients

Local (stdio)

Client	Status	Notes
Claude Desktop	Supported	Full support
Continue.dev	Supported	VS Code extension
Cursor	Planned	IDE integration

Remote (HTTP/SSE)

Client	Status	Notes
Claude.ai	Supported	Via MCP discovery
Custom Apps	Supported	Direct API calls

Claude.ai Integration

For remote ByteMCP servers, Claude.ai can discover and use MCP tools.

MCP Discovery

ByteMCP exposes discovery at:

https://mcp.bytefederal.com/.well-known/mcp-server

Response:

{
  "name": "bytefederal-bitcoin-atm",
  "version": "1.0.0",
  "endpoints": {
    "sse": "https://mcp.bytefederal.com/sse",
    "http": "https://mcp.bytefederal.com/mcp"
  },
  "capabilities": ["tools"],
  "tools": [
    "find_nearest_bitcoin_atm",
    "get_atm_details",
    "check_atm_status",
    "list_atms_by_city"
  ],
  "supported_transports": ["sse"],
  "auth_required": false
}

Website Integration

Add to your website’s robots.txt:

User-agent: Claude-Web
Allow: /.well-known/

MCP-Server: https://mcp.bytefederal.com/.well-known/mcp-server

Continue.dev (VS Code)

Configuration

Add to your Continue config:

{
  "mcpServers": [
    {
      "name": "bytefederal-atm",
      "transport": {
        "type": "stdio",
        "command": "node",
        "args": ["/path/to/byte-mcp/dist/index.js"],
        "env": {
          "DB_HOST": "localhost",
          "DB_DATABASE": "bytefederal",
          "DB_USER": "user",
          "DB_PASSWORD": "password"
        }
      }
    }
  ]
}

Custom Applications

Direct HTTP Calls

For custom integrations, call the HTTP endpoint directly:

async function findNearestATM(location) {
  const response = await fetch('https://mcp.bytefederal.com/mcp', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
      jsonrpc: '2.0',
      id: 1,
      method: 'tools/call',
      params: {
        name: 'find_nearest_bitcoin_atm',
        arguments: { location, limit: 5 }
      }
    })
  });

  return response.json();
}

SSE Connection

For real-time updates:

const eventSource = new EventSource('https://mcp.bytefederal.com/sse');

eventSource.onmessage = (event) => {
  const data = JSON.parse(event.data);
  handleMCPMessage(data);
};

function sendToolCall(toolName, args) {
  fetch('https://mcp.bytefederal.com/sse', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
      jsonrpc: '2.0',
      id: Date.now(),
      method: 'tools/call',
      params: { name: toolName, arguments: args }
    })
  });
}

LangChain Integration

from langchain.tools import Tool
import requests

def find_atm(query: str) -> str:
    response = requests.post(
        'https://mcp.bytefederal.com/mcp',
        json={
            'jsonrpc': '2.0',
            'id': 1,
            'method': 'tools/call',
            'params': {
                'name': 'find_nearest_bitcoin_atm',
                'arguments': {'location': query}
            }
        }
    )
    return response.json()['result']

atm_tool = Tool(
    name="find_bitcoin_atm",
    description="Find Bitcoin ATMs near a location",
    func=find_atm
)

OpenAI Function Calling

Convert MCP tools to OpenAI function format:

const functions = [
  {
    name: "find_nearest_bitcoin_atm",
    description: "Find Bitcoin ATMs near a location",
    parameters: {
      type: "object",
      properties: {
        location: {
          type: "string",
          description: "Address, city/state, or ZIP code"
        },
        limit: {
          type: "number",
          description: "Maximum results (1-50)"
        }
      },
      required: ["location"]
    }
  }
];

// Bridge function calls to ByteMCP
async function handleFunctionCall(name, args) {
  return fetch('https://mcp.bytefederal.com/mcp', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
      jsonrpc: '2.0',
      id: 1,
      method: 'tools/call',
      params: { name, arguments: args }
    })
  }).then(r => r.json());
}

Authentication

ByteMCP currently operates without authentication for public ATM data. For restricted deployments:

1. Add API key middleware
2. Use OAuth for user-specific queries
3. Implement rate limiting per client

CORS Considerations

Remote ByteMCP allows:

• *.anthropic.com
• *.claude.ai
• bytefederal.com

Add your domain in remoteServer.ts for custom integrations.

Next Steps

• API Reference - Full API documentation
• Remote Deployment - Set up remote server