Credits

🚀 Kafka MCP Server

The Kafka MCP Server is an implementation of the Model Context Protocol (MCP). It offers integration with Apache Kafka, enabling interactions with Kafka clusters through natural language (LLMs).

demo

🚀 Quick Start

Prerequisites

You need Docker or Golang to run the MCP Server locally.

Accessing a Kafka Cluster

You need access to a Kafka cluster. Follow the quickstart documentation:

docker pull apache/kafka:4.0.0
docker run -p 9092:9092 apache/kafka:4.0.0

Now, Kafka can be accessed via localhost:9092.

Using Claude Desktop

Docker:

{
  "mcpServers": {
    "kafka": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "KAFKA_MCP_BOOTSTRAP_SERVERS",
        "ghcr.io/cefboud/kafka-mcp-server"
      ],
      "env": {
        "KAFKA_MCP_BOOTSTRAP_SERVERS": "localhost:9092"
      }
    }
  }
}

Local Build

cd <Working Directory>
git clone https://github.com/CefBoud/kafka-mcp-server.git
cd kafka-mcp-server
go build -o kafka-mcp-server cmd/kafka-mcp-server/main.go

{
  "mcpServers": {
    "kafka": {
      "command": "<Working Directory>/kafka-mcp-server/kafka-mcp-server",
      "args": [
        "stdio"
      ],
      "env": {
        "KAFKA_MCP_BOOTSTRAP_SERVERS": "localhost:9092"
      }
    }
  }
}

Options

      --bootstrap-servers string   Comma-separated list of Kafka servers to connect to.
      --enable-command-logging     When enabled, the server will log all command requests and responses to a log file.
      --log-file string            Path to the log file.
      --read-only                  Restrict the server to perform only read-only operations.

All options can be passed as environment variables, in uppercase, replacing hyphens with underscores and prefixing with MCP_KAFKA_. For example, --bootstrap-servers becomes MCP_KAFKA_BOOTSTRAP_SERVERS.

✨ Features

Available MCP Tools

• [x] List topics
• [x] Create topics
• [x] Consume messages.
• [x] Produce messages.
• [x] Describe the cluster (list of brokers and controllers)
• [x] List consumer groups and their lags.
• [x] Get the earliest and latest offsets of a topic (GetOffsetShell)
• [ ] Reset consumer group offsets.
• [ ] Kafka Connect ??
• [ ] Schema Registry ??

🔀 MultiplexTool

demo MultiplexTool

Running multiple sequential tools, especially when each tool depends on the output of the previous one, can be time-consuming and cumbersome. This requires multiple round-trips between the client and the server.

MultiplexTool solves this problem by allowing the client to batch a list of tool calls into one request and execute them sequentially. It supports dynamic dependencies between tools, allowing you to reference earlier outputs using prompt-based placeholders.

If the tool input depends on previous results, the client can generate the input via a prompt using the PROMPT_ARGUMENT: format (currently only supports Gemini). Example:
"userId": "PROMPT_ARGUMENT: ID of the created user"

CLI Options:

• --enable-multiplex Enable the multiplexing feature
• --multiplex-config string Specify the path to the multiplexing configuration file

Configuration Example

{
  "mcpServers": {
    "kafka": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "KAFKA_MCP_BOOTSTRAP_SERVERS",
        "ghcr.io/cefboud/kafka-multiplex-server:latest"
      ],
      "env": {
        "KAFKA_MCP_BOOTSTRAP_SERVERS": "localhost:9092",
        "GEMINI_API_KEY": "....."
      }
    }
  }
}

📄 License

• github.com/github/github-mcp-server
• github.com/mark3labs/mcp-go
• github.com/IBM/sarama