Drop Thaddeus into Claude and get a Senior Kafka Streaming Engineer who states the ordering, retention, and delivery guarantee a topic actually needs before it gets created.
Thaddeus covers the full event-driven build cycle on Apache Kafka and Confluent: partition key and count sizing tied to real throughput and ordering needs, schema registry setup with an explicit compatibility mode, stream processing topology in Kafka Streams or ksqlDB with real windowing, event sourcing pattern design with a validated replay contract, consumer group scaling and rebalancing tuning, and exactly-once semantics reserved for cases that genuinely require it.
What you get
-
→Partition key chosen for the actual ordering guarantee required, not maximum parallelism
-
→Partition count sized from real throughput math, not a round number
-
→Schema Registry compatibility mode stated explicitly for every topic
-
→Exactly-once semantics reserved for cases where duplicate processing causes real harm
📄 thaddeus-kafka-streaming-engineer.skill
Under 2 min install
Works with Claude, ChatGPT & any AI chat
How to install
Download the .skill package → open Claude → paste SKILL.md into your Project Instructions or system prompt → describe your requirement → Thaddeus builds the answer. Includes a full worked example so you see exactly what you get.