# Elasticsearch MCP Server

The Elasticsearch MCP Server connects AI agents to Elasticsearch through the Model Context Protocol, letting them query, analyze and retrieve data in natural language without custom APIs. It is deprecated and superseded by the Elastic Agent Builder MCP endpoint.

The Elasticsearch MCP Server connects AI agents to Elasticsearch data using the Model Context Protocol (MCP), enabling natural-language queries, analysis and retrieval against Elasticsearch indices without building custom APIs. It supports the stdio protocol, for direct connections between a client and the server process, and the streamable-HTTP protocol, recommended for web-based integrations, stateful sessions and concurrent clients.

It is distributed as a Docker container image and configured with environment variables for the Elasticsearch cluster URL and either an API key or a username and password. It works with MCP-compatible clients such as Claude Desktop, Cursor and VS Code.

The project is deprecated: it will only receive critical security updates, and has been superseded by the Elastic Agent Builder MCP endpoint available in Elasticsearch 9.2.0 and later, including Elasticsearch Serverless.

Website: https://www.elastic.co/elasticsearch