DataBridgeAI MCP Client

A Node.js client that bridges MCP (Model Context Protocol) applications like Claude Desktop to the DataBridgeAI HTTP service.

Installation

# Install globally
npm install -g databridgeai-mcp-client

# Or use with npx (recommended)
npx databridgeai-mcp-client

Usage

With Claude Desktop

Add this configuration to your Claude Desktop MCP settings:

{
  "mcpServers": {
    "databridgeai": {
      "command": "npx",
      "args": ["-y", "databridgeai-mcp-client"],
      "env": {
        "MCP_SERVER_URL": "http://localhost:8000/api/mcp",
        "DATABRIDGEAI_AUTH_TOKEN": "your-auth-token-here"
      }
    }
  }
}

Environment Variables

• MCP_SERVER_URL: URL of the DataBridgeAI MCP HTTP service (default: http://localhost:8000/api/mcp)
• MCP_TIMEOUT: Request timeout in milliseconds (default: 30000)
• DATABRIDGEAI_AUTH_TOKEN: Bearer token for DataBridgeAI authentication (required)

Command Line Options

# Test connection
npx databridgeai-mcp-client --test

# Show help
npx databridgeai-mcp-client --help

# Show version
npx databridgeai-mcp-client --version

# Use custom URL
npx databridgeai-mcp-client http://localhost:8000/api/mcp

Getting Your Auth Token

1. Log into your DataBridgeAI frontend application
2. Open browser developer tools (F12)
3. Go to the Network tab
4. Make any API request
5. Look for the Authorization header in the request
6. Copy the token part after Bearer

Development

# Clone the repository
git clone https://github.com/databridgeai/databridgeai-mcp-client.git
cd databridgeai-mcp-client

# Test the client
node index.js --test

Troubleshooting

Connection Issues

• Make sure your DataBridgeAI backend is running on the correct port (default: 8000)
• Verify your auth token is valid and not expired
• Check that the MCP endpoint is accessible at /api/mcp

Authentication Issues

• Ensure DATABRIDGEAI_AUTH_TOKEN is set correctly
• The token should be a valid JWT Bearer token
• Check that your user account has the necessary permissions

License

MIT