mcp-server-opensearch: An OpenSearch MCP Server

What is mcp-server-opensearch?

mcp-server-opensearch is a Model Context Protocol server that integrates with OpenSearch, a distributed search and analytics engine, to act as a semantic memory layer for LLM applications. It provides a standardized way to store and retrieve memories in an OpenSearch database.

How to use mcp-server-opensearch?

Install via Smithery (npx -y @smithery/cli install @ibrooksSDX/mcp-server-opensearch --client claude) or run directly with uv run mcp-server-opensearch --opensearch-url "http://localhost:9200" --index-name "my_index". Configure using command-line arguments or environment variables (OPENSEARCH_HOST, OPENSEARCH_HOSTPORT, INDEX_NAME). Add to Claude Desktop's claude_desktop_config.json with the appropriate command and args.

Key features of mcp-server-opensearch

• Provides a single tool search-openSearch to query OpenSearch.
• Acts as a semantic memory layer on top of OpenSearch.
• Supports configuration via CLI arguments or environment variables.
• Can be installed automatically via Smithery for Claude Desktop.
• Uses FastMCP for development and testing.
• Includes a demo script (demo.py) for quick setup.

Use cases of mcp-server-opensearch

• Storing and retrieving memories in an OpenSearch database from an LLM.
• Enhancing AI chat interfaces with persistent searchable context.
• Building custom AI workflows that require a semantic memory layer.

FAQ from mcp-server-opensearch

What is the current blocker for the async client?

The async client from OpenSearch