Kafka MCP Server
A comprehensive MCP server for Apache Kafka operations, enabling seamless integration with Claude Desktop and other MCP clients for cluster management, topic operations, and message streaming.
README
Kafka MCP Server
A comprehensive Model Context Protocol (MCP) server for Apache Kafka operations, enabling seamless integration with Claude Desktop and other MCP clients.
🚀 Features
- Cluster Management: Monitor cluster health, broker information, and metadata
- Topic Operations: Create, list, describe, and delete Kafka topics
- Message Operations: Send and consume messages with flexible configuration
- Consumer Group Management: List and describe consumer groups
- Real-time Monitoring: Get topic metrics, offsets, and performance data
- Health Checks: Comprehensive cluster health monitoring
- Secure Configuration: Support for SASL, SSL, and various authentication methods
📋 Prerequisites
- Python 3.8+
- Apache Kafka cluster (local or remote)
- pip package manager
🛠️ Installation
-
Clone the repository:
git clone https://github.com/aswinayyolath/kafka-mcp-server.git cd kafka-mcp-server -
Create a virtual environment:
python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate -
Install dependencies:
pip install mcp kafka-python -
Configure environment variables:
cp .env.template .env # Edit .env with your Kafka configuration
⚙️ Configuration
Environment Variables
Create a .env file based on .env.template:
# Basic Kafka Configuration
KAFKA_BOOTSTRAP_SERVERS=localhost:9092
KAFKA_SECURITY_PROTOCOL=PLAINTEXT
# SASL Configuration (if needed)
KAFKA_SASL_MECHANISM=PLAIN
KAFKA_SASL_USERNAME=your_username
KAFKA_SASL_PASSWORD=your_password
# SSL Configuration (if needed)
KAFKA_SSL_CAFILE=/path/to/ca.pem
KAFKA_SSL_CERTFILE=/path/to/cert.pem
KAFKA_SSL_KEYFILE=/path/to/key.pem
Docker Setup (Optional)
Start a local Kafka cluster using Docker:
docker-compose up -d
This will start Kafka on localhost:9092.
🧪 Testing
Run the comprehensive test suite to validate your setup:
# Set environment variables
export KAFKA_BOOTSTRAP_SERVERS=localhost:9092
export KAFKA_SECURITY_PROTOCOL=PLAINTEXT
# Run tests
python test_kafka_mcp.py
The test suite validates:
- Kafka connectivity
- MCP server startup
- All available tools
- Topic operations
- Message operations
- Resources and prompts
🔧 Usage
Standalone Server
Run the MCP server directly:
python kafka_mcp_server.py
Claude Desktop Integration
Add to your Claude Desktop configuration (claude_desktop_config.json):
{
"mcpServers": {
"kafka": {
"command": "python",
"args": ["/path/to/kafka_mcp_server.py"],
"env": {
"KAFKA_BOOTSTRAP_SERVERS": "localhost:9092",
"KAFKA_SECURITY_PROTOCOL": "PLAINTEXT"
}
}
}
}
🛠️ Available Tools
Cluster Operations
get_cluster_info- Get comprehensive cluster informationhealth_check- Perform cluster health check
Topic Operations
list_topics- List all topics in the clustercreate_topic- Create a new topicdescribe_topic- Get detailed topic informationdelete_topic- Delete a topicget_topic_metrics- Get comprehensive topic metricsget_topic_offsets- Get partition offsets
Message Operations
send_message- Send a single messagesend_batch_messages- Send multiple messagesconsume_messages- Consume messages from a topic
Consumer Group Operations
list_consumer_groups- List all consumer groupsdescribe_consumer_group- Get detailed consumer group information
📚 Resources
The server provides MCP resources for easy access to cluster information:
kafka://cluster/info- Real-time cluster informationkafka://topics/list- Current topics list
🎯 Prompts
Built-in prompts for common scenarios:
kafka_monitoring_prompt- Generate monitoring and troubleshooting guidancekafka_troubleshooting_prompt- Get help with specific issues
💡 Example Usage with Claude
Once integrated with Claude Desktop, you can ask:
- "Can you check the health of my Kafka cluster?"
- "List all topics and their partition counts"
- "Create a topic called 'user-events' with 3 partitions"
- "Send a test message to the user-events topic"
- "Show me the last 10 messages from user-events"
- "What consumer groups are active?"
🔒 Security
Authentication
The server supports various Kafka authentication methods:
- PLAINTEXT: No authentication (development only)
- SASL_PLAINTEXT: SASL authentication over plain connection
- SASL_SSL: SASL authentication over SSL
- SSL: SSL client certificate authentication
Best Practices
- Never commit
.envfiles - Use.env.templatefor examples - Use SSL in production - Always encrypt connections to production clusters
- Limit permissions - Use dedicated service accounts with minimal required permissions
- Monitor access - Log and monitor MCP server usage
🐛 Troubleshooting
Common Issues
-
Connection refused:
- Verify Kafka is running
- Check
KAFKA_BOOTSTRAP_SERVERSconfiguration - Ensure network connectivity
-
Authentication failures:
- Verify SASL credentials
- Check SSL certificate paths
- Validate security protocol settings
-
Topic not found:
- Ensure topic exists
- Check topic name spelling
- Verify permissions
Debug Mode
Enable debug logging by setting:
export PYTHONPATH=.
python -c "import logging; logging.basicConfig(level=logging.DEBUG)"
python kafka_mcp_server.py
🤝 Contributing
- Fork the repository
- Create a feature branch:
git checkout -b feature-name - Make your changes
- Run tests:
python test_kafka_mcp.py - Commit changes:
git commit -am 'Add feature' - Push to branch:
git push origin feature-name - Submit a pull request
📄 License
This project is licensed under the Apache License 2.0 - see the LICENSE file for details.
🙏 Acknowledgments
- Model Context Protocol - For the MCP specification
- Apache Kafka - For the distributed streaming platform
- kafka-python - For the Python Kafka client
📞 Support
For issues and questions:
- Check the troubleshooting section
- Run the test suite to validate your setup
- Review Kafka and MCP documentation
- Open an issue with detailed error information
Happy Kafka streaming with MCP! 🎉
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。