I make CockroachDB a natural fit wherever people already build, from AI coding tools and agent frameworks to change data capture, orchestration, and the mainframe. Most of my work ships as open source in the projects developers actually use.
CockroachDB now shows up as a first class option inside the major AI coding tools. I built and maintain the official plugins, all sharing one skills library and MCP backend, so an AI agent can stand up a local cluster and build an app that follows CockroachDB best practices by default.
Schema aware assistance and query generation in Copilot.
cockroachdb/copilot-plugin33+ agent skills across 9 operational domains, shared by every plugin. Install once, use across 40+ agents.
cockroachlabs/cockroachdb-skillsModel Context Protocol backend that connects agents to CockroachDB.
cockroachdb/cockroachdb-mcp-serverThe highest leverage way to grow an ecosystem is to meet developers where they are. These landed upstream, in the tools teams already depend on.
Automatic retry for transient errors, with no application code changes.
viragtripathi/cockroachdb-jdbc-wrapperMulti-region resilience demo you can run yourself.
viragtripathi/cockroach-chaos-demoReference for running Apache Airflow on CockroachDB.
viragtripathi/cockroachdb-airflow-pocRunnable change data capture configurations.
viragtripathi/debezium-cockroachdb-examplesRecent pieces cover agentic AI on watsonx with CockroachDB, running CockroachDB on IBM LinuxONE, multi source reasoning with the MCP Toolbox, hybrid vector search, and durable agents with LangChain and LangGraph.