VoltDB Adds Agent Skills and MCP Server for AI-Assisted Development

Summary

Volt Active Data documents two ways to connect its database software with AI agents and chatbots: Agent Skills that teach coding agents about VoltDB products, and an MCP Server that provides controlled access to a running VoltDB database. The documentation says skills can improve response accuracy and reduce the AI credits needed by narrowing the agent’s focus. Volt provides three initial MIT-licensed skills in a public GitHub repository: voltdb-development for client applications, schemas, stored procedures, and integration tests; voltdb-kubernetes for Helm charts and Kubernetes deployment scripts; and voltsp for Volt Stream Processing pipelines in Java or YAML. Installation depends on the agent or LLM vendor; the Claude Code example copies selected skill directories into ~/.claude/skills. Agents may apply installed skills automatically based on a request, while users can explicitly invoke one with a slash command such as /voltdb-development. An explicitly invoked skill remains active for the current Claude Code session and must be invoked again after a restart. The chapter illustrates this workflow with a real-time DDoS detection application request. It presents the MCP Server as the complementary runtime capability: skills provide knowledge about VoltDB, while the server lets an agent query and analyze a specific database’s contents, status, and performance.