Practical overview
When it fits
Use it when an existing MCP client needs agent access to an Elasticsearch cluster and you can run the Docker container with valid credentials and network connectivity. Consider the stated successor for new work because this server is deprecated.
Who it is for
- Teams connecting MCP-compatible AI clients to an Elasticsearch 8.x or 9.x cluster.
- Developers choosing between direct stdio connections and streamable-HTTP integrations.
- Operators who can provide Docker, Elasticsearch credentials and network access to the cluster.
Documented uses
- Let an AI agent query Elasticsearch indices without building a custom API for the connection.
- Use indexed Elasticsearch data as an input to an agent application.
- Choose a direct stdio or streamable-HTTP transport for an MCP client.
Requirements & limits
- Requires a reachable Elasticsearch 8.x or 9.x cluster, authentication credentials, Docker and an MCP client.
- Stdio supports direct client-to-server process communication; streamable HTTP is documented for web integrations and concurrent clients.
- The repository says the project receives only critical security updates and names Elastic Agent Builder's MCP endpoint as its successor.
- The source repository is Apache-2.0 licensed and identifies Rust as its primary language.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Configure a direct stdio connection
Run the container in stdio mode and configure a client such as Claude Desktop to connect to it.
Before you start
- An Elasticsearch 8.x or 9.x cluster reachable from the Docker environment
- An Elasticsearch API key, or a username and password
- Docker installed and running
- An MCP client that supports stdio
- Network connectivity between the server environment and Elasticsearch
Steps
Set ES_URL and either ES_API_KEY or ES_USERNAME and ES_PASSWORD for the container.
Run the Elasticsearch MCP Server Docker image with the stdio argument and interactive input enabled.
Add the documented Docker command and Elasticsearch connection values to the MCP client's server configuration.
Expected outcome
The MCP client has a configured stdio connection to the server for agent access to Elasticsearch indices.
Guide limitations
- The README's instructions apply to version 0.4.0 and later.
- The project is deprecated and will receive only critical security updates.
- The README's connection-verification steps are separate; this recipe does not claim a connection was tested.
Run the server over streamable HTTP
Start the container in HTTP mode for a web integration or multiple concurrent clients.
Before you start
- An Elasticsearch 8.x or 9.x cluster reachable from the Docker environment
- An Elasticsearch API key, or a username and password
- Docker installed and running
- An MCP client configured for streamable HTTP
- Network connectivity between the server environment and Elasticsearch
Steps
Set ES_URL and either ES_API_KEY or ES_USERNAME and ES_PASSWORD.
Run the Elasticsearch MCP Server Docker image with the http argument and expose port 8080.
Configure the client to use the server's /mcp endpoint; use the documented proxy approach if the client supports only stdio.
Expected outcome
The server provides a streamable-HTTP MCP endpoint at /mcp; the README also identifies /ping as a health-check endpoint.
Guide limitations
- The project marks SSE as deprecated and recommends streamable HTTP instead.
- The README's instructions apply to version 0.4.0 and later.
- The project is deprecated and will receive only critical security updates.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 13 October 2026
- Original website title
- elastic/mcp-server-elasticsearch · Official homepage title · retained in its source language
Related options
Elastic Agent Builder MCP endpoint — The repository identifies this as the successor; it is available in Elastic 9.2.0+ and Elasticsearch Serverless projects.