Happy slice of mcp pizzaMCP.pizza

mcp-confluent

MCP.Pizza Chef: confluentinc

Seventy-three tools cover Confluent Cloud, Confluent Platform and plain Apache Kafka: listing topics, posting and reading messages, checking how far a consumer has fallen behind, running Flink SQL, managing connectors and searching Confluent's own documentation. The documentation and diagnostic tools need no credentials at all; anything touching a real cluster needs Confluent keys or a browser sign-in. Deleting topics, registered message formats and connectors is included with no undo step, so fence the assistant in with the allow and block lists.

Coding
Data

Use This MCP server To

Ask which topics exist and how each one is configured Find out why a consumer group is falling behind Post a test message to a topic and read it back Look up a Confluent documentation answer without leaving the chat Work out why a connector keeps failing Run a Flink SQL statement and see the results

README

Confluent MCP Server

npm version License: MIT

An open-source MCP server that enables AI assistants to interact with Confluent Cloud, Confluent Platform, and standalone Apache Kafka deployments through natural language. It provides 50+ tools across Kafka, Flink SQL, Schema Registry, Connectors, Tableflow, and more -- usable from any MCP-compatible client including Claude Desktop, Claude Code, Cursor, VS Code, Goose, and Gemini CLI.

Tip

Already a Confluent Cloud customer? Confluent offers a fully managed MCP server with no local server to run and no dependencies to install. It provides access to your Confluent Cloud resources with AI-powered connector diagnostics, governed by your existing RBAC permissions. Use this open-source server if you need Confluent Platform / self-managed Kafka support, or want to customize and extend the toolset.

NOTE: The open-source MCP server is a community-supported project. Confluent doesn’t provide dedicated support for it, and support is best-effort only, with no service level commitments. If you run into an issue or want to contribute, open an issue or pull request directly in this repository.

Quick Start

Prerequisites: Node.js 22.19.0+. If you want to interact with Confluent Cloud, you need to create an account first.

  1. Generate a quick config.yaml file in your project root:
npx @confluentinc/mcp-confluent --init-config
  1. Edit the config.yaml file with your connection details, then:
npx @confluentinc/mcp-confluent --config ./config.yaml

See Getting Started for full setup instructions and Configuring MCP Clients for integration with your preferred AI tool.

Table of Contents

mcp-confluent FAQ

How much can it really do?
Seventy-three tools are registered, rather more than the fifty-plus advertised, but only those matching your configuration switch on. The rest stay hidden and can explain why on request.
Can it delete things?
Yes. Topics, registered message formats, connectors, Flink statements and tags can all be removed, and there is no overall read-only switch. Limit it with the allow-tools or block-tools lists.
Do I need a key?
For documentation and diagnostics, no. For anything touching a cluster, yes, though several areas accept a browser sign-in instead of hand-made keys.
Does it report anything about my usage?
Yes. Usage data and error reports are sent by default; setting DO_NOT_TRACK to true in your environment turns both off.
How hard is setup?
You need a recent Node version, one command that writes a starter config file, your connection details filled in, then the server pasted into your assistant's settings.
Which apps does it work with?
Claude Desktop, Claude Code, Cursor, VS Code, Goose and Gemini CLI are all named, plus any other MCP client. Confluent Cloud customers can use Confluent's own hosted version instead.
Is it officially supported?
It is open source and community supported. Confluent offers only best-effort help on it and makes no service commitments.