Connects Claude to Kafka clusters for DataOps monitoring and troubleshooting. Exposes four tools: check consumer group lag with diagnostic patterns, inspect topic configuration and ISR health, view broker status and under-replicated partitions, and list topics. The diagnostics aren't generic, they're based on real production incidents like detecting Flink crashes from inactive consumers plus growing lag, or flagging ClusterIdMismatch from Docker volume conflicts. Reach for this when you're debugging CDC pipelines or need Claude to assess cluster health and consumer lag issues. Configuration is straightforward, just point it at your bootstrap servers via environment variables.
claude mcp add --transport stdio io.github.aguantar-kafka-dataops-mcp -- uvx kafka-dataops-mcp