Overview MCP Server

What is mcp-kafka?

mcp-kafka is a server-side implementation of the Model Context Protocol (MCP) for Apache Kafka. It enables language models (LLMs/SLMs) to interact with Kafka and its ecosystem, including Kafka Connect, Burrow, and Cruise Control.

How to use mcp-kafka?

Configure the server by setting environment variables for the APIs you want to enable (Kafka bootstrap servers, Kafka Connect URL, Burrow URL, Cruise Control URL). Then plug it into any MCP client such as Claude Desktop, Cursor, Windsurf, Langchain MCP adapter, or Azure OpenAI. Publishing to PyPI and Docker is planned but not yet available.

Key features of mcp-kafka

• Tools for Kafka: consume, produce, describe clusters, topics, consumer groups, delegation tokens, log dirs, configs, ACLs.
• Tools for Kafka Connect: cluster info, config, connectors, connector plugins, loggers.
• Tools for Burrow: health check, list clusters, describe cluster, list consumers with group detail, list topics, check consumer group status.
• Tools for Cruise Control: get state, cluster load, partition resource utilization, partition/replica state, optimization proposals, user request result.
• Resources exposed: topics, connectors, consumer groups.
• Support for multiple MCP client integrations.

Use cases of mcp-kafka

• Monitor consumer group lag and health via Burrow.
• Rebalance Kafka clusters and view optimization proposals using Cruise Control.
• Manage and inspect Kafka connectors programmatically.
• Produce and consume messages to/from Kafka topics through an LLM.
• Retrieve metadata about topics, consumer groups, and cluster configuration.