MXCP

Name: MXCP
Availability: InStock
Author: raw-labs

Supports UI

by raw-labs

Enterprise-grade framework for building production-ready MCP servers with security and data quality.

0 stars

1 views

Works in:claude

Exposes:ToolsResourcesPrompts

View on GitHub Docs

What it does

MXCP is a comprehensive methodology and framework for building production-ready MCP servers. It emphasizes a "data-modeling first" approach, integrating dbt for data quality and providing enterprise-grade security, auditing, and type safety for AI-to-data infrastructure.

Tools

MXCP is a framework for creating tools rather than a single-purpose server. It allows developers to implement:

SQL-based tools: High-performance data queries using DuckDB.
Python-based tools: Complex business logic, ML models, and external API integrations.
Policy-driven access: RBAC and data filtering to ensure AI safety.

Installation

Install the framework via pip:

pip install mxcp

Then initialize a project:

mxcp init --bootstrap

To connect to Claude Desktop:

{
  "mcpServers": {
    "my-tools": {
      "command": "mxcp",
      "args": ["serve", "--transport", "stdio"],
      "cwd": "/path/to/my-ai-tools"
    }
  }
}