NoCodeBackend MCP Server

Official MCP (Model Context Protocol) server for NoCodeBackend. Enables native integration with AI coding tools like Cursor, Windsurf, Claude Desktop, and others.

Installation

``bash
npm install -g @nocodebackend/mcp
`

Or run directly with npx:

`bash
npx @nocodebackend/mcp
`

Quick Start

$3

Get your API token from the NoCodeBackend dashboard: Settings → API Token

$3

Add to your Cursor MCP settings (~/.cursor/mcp.json):

`json
{
"mcpServers": {
"nocodebackend": {
"command": "npx",
"args": ["@nocodebackend/mcp"]
}
}
}
`

$3

When you first use the MCP, you'll only see the login tool. Use it to authenticate:

`
Use tool: login
Token: your_api_token_here
`

After successful login, all tools become available.

Alternative: Environment Variable

Instead of using the login tool, you can set your token via environment variable:

`json
{
"mcpServers": {
"nocodebackend": {
"command": "npx",
"args": ["@nocodebackend/mcp"],
"env": {
"NCB_TOKEN": "your_api_token"
}
}
}
}
`

Available Tools

| Tool | Description |
|------|-------------|
| login | Authenticate with your NoCodeBackend API token |
| create_database | Create a new database with tables |
| list_databases | List all your databases |
| get_schema | Get the schema of a database |
| get_integration_guide | Get integration rules and best practices |
| get_integration_prompts | Get proxy code for your Next.js app |
| create_secret_key | Create API keys for your database |
| list_secret_keys | List API keys for a database |
| execute_sql | Execute SQL statements |

Example Usage

Once configured and logged in, you can ask your AI assistant:

- "Create a new database called my_app with users and posts tables"
- "Get the integration guide for NoCodeBackend"
- "Show me the schema of my database"
- "Create a secret key for my production app"

Resources

The server exposes your databases as MCP resources:

- nocodebackend://database/{name}` - Database schema as JSON

Support

- Documentation: nocodebackend.com/docs
- Issues: GitHub Issues

License

MIT