mcp-nats

mcp-nats

7

The mcp-nats project is an MCP server designed for integration with the NATS messaging system, allowing AI models to interact through a standardized protocol. It offers read-only operations for server management, stream operations, key-value management, and message publishing, supporting secure multi-account handling.

Overview

This project provides a Model Context Protocol (MCP) server for NATS, enabling AI models to interact with NATS messaging systems via a standardized interface. It offers tools for NATS server interactions, suitable for AI applications that work with messaging systems.

Features

  • Server Management (Read-only Operations): Inspect and monitor NATS servers, health checks, server info retrieval.
  • Stream Operations (Read-only Operations): View and inspect NATS streams, query stream state and info, retrieve messages.
  • Key-Value Operations: Manage KV buckets, store/retrieve key-value pairs, watch for updates, delete keys and buckets.
  • Publish Operations: Publish messages to NATS subjects, support for various message formats, asynchronous publishing.
  • Multi-Account Support: Handle multiple NATS accounts, secure credential management.
  • MCP Integration: Implements MCP server specification, compatible with clients like Claude Desktop, provides standardized tool definitions for LLM interaction, supports safe, read-only AI interactions with NATS.

Requirements

  • Go 1.24 or later
  • NATS server via URL
  • NATS credentials
  • MCP-compatible client (e.g., Claude Desktop)