OpenSearch MCP <> LLM Integration Guide

@madhankb

リポジトリ

基本情報

カテゴリ

データベース

トランスポート

stdio

作者

madhankb

OpenSearch MCP <> LLM Integration Guide について

Step by step guide for integrating LLMs with OpenSearch MCP servers

OpenSearch MCP <> LLM Integration Guide の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "opensearch-mcp-integration": {
      "command": "docker",
      "args": [
        "compose",
        "up",
        "-d"
      ]
    }
  }
}

OpenSearch MCP <> LLM Integration Guide のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is OpenSearch MCP <> LLM Integration Guide?

This guide provides step-by-step instructions for integrating Amazon Q and Anthropic Claude Desktop with OpenSearch using MCP servers, enabling natural language interactions with an OpenSearch cluster. It covers setting up a local OpenSearch instance, installing the opensearch-mcp-server, configuring LLM clients, and using natural language commands to manage indices and documents.

How to use OpenSearch MCP <> LLM Integration Guide?

Install the MCP server via pipx install opensearch-mcp-server. Configure your LLM client by creating a JSON file (mcp.json for Amazon Q or claude_desktop_config.json for Claude Desktop) with the command path and environment variables (OPENSEARCH_URL, OPENSEARCH_SSL_VERIFY, OPENSEARCH_USERNAME, OPENSEARCH_PASSWORD). Start a local OpenSearch instance using the provided docker-compose.yml, then use natural language prompts in the LLM to perform operations like creating indices, adding documents, and searching.

Key features of OpenSearch MCP <> LLM Integration Guide

Provides MCP tools for managing OpenSearch indices and documents.
Supports creating, reading, updating, and deleting indices and documents.
Enables search and delete by query using natural language.
Offers cluster health and statistics inspection.
Includes alias management and a general-purpose API request tool.
Integrates with Amazon Q and Anthropic Claude Desktop.

Use cases of OpenSearch MCP <> LLM Integration Guide

Create a new index with specified fields using natural language.
Add, retrieve, update, and delete documents without writing raw Elasticsearch queries.
Search for documents using plain‑language descriptions.
Get index mappings and cluster health information on demand.
Perform administrative tasks like alias management and custom API requests.

よくある質問

What are the prerequisites for using this setup?

You need Docker and Docker Compose, Python 3.x, pipx, and the LLM clients (Amazon Q CLI or Anthropic Claude Desktop).

How do I start a local OpenSearch instance?

Create a `docker-compose.yml` with the provided configuration and run `docker compose up -d`. The instance will be available on ports 9200 and 9600.

What MCP tools are available?

The server exposes tools including `list_indices`, `get_index`, `create_index`, `delete_index`, `search_documents`, `index_document`, `get_document`, `delete_document`, `delete_by_query`, `get_cluster_health`, `get_cluster_stats`, `list_aliases`, `get_alias`, `put_alias`, `delete_alias`, and `general_api_request`.

How do I configure Amazon Q or Claude Desktop to use the MCP server?

Create a JSON configuration file (e.g., `~/.aws/amazonq/mcp.json` for Amazon Q or the Claude Desktop config path) with the `opensearch-mcp-server` command and environment variables for the OpenSearch URL, credentials, and SSL settings.

What security precautions are recommended?

Use strong passwords in production, enable SSL, and regularly rotate credentials. The guide defaults to SSL verification disabled for local testing, but this should be removed in production.

「データベース」の他のコンテンツ

MotherDuck's DuckDB MCP Server

motherduckdb

Local MCP server for DuckDB and MotherDuck

498

MCP Alchemy

runekaagaard

A MCP (model context protocol) server that gives the LLM access to and knowledge about relational databases like SQLite, Postgresql, MySQL & MariaDB, Oracle, and MS-SQL.

412