OpenSearch MCP LLM Integration Guide

This guide provides step-by-step instructions for integrating Amazon Q & Anthropic Claude desktop with OpenSearch using MCP servers for natural language interactions.

Anthropic Claude

Amazon Q

Prerequisites

- Docker and Docker Compose installed
- Python 3.x installed
- pipx installed (pip install pipx)
- Amazon Q CLI
- Anthropic Claude Desktop

1. Setting up Local OpenSearch Instance

1. Create a new directory for your project:

   mkdir opensearch-mcp && cd opensearch-mcp

   

2. Create a docker-compose.yml file with the following content:

   version: '3'

   services:

     opensearch:

       image: opensearchproject/opensearch:latest

       environment:

         - discovery.type=single-node

         - bootstrap.memory_lock=true

         - "OPENSEARCH_JAVA_OPTS=-Xms512m -Xmx512m"

         - "OPENSEARCH_INITIAL_ADMIN_PASSWORD=myStrongPassword123!"

         - "plugins.security.ssl.http.enabled=true"

         - "plugins.security.ssl.transport.enabled=true"

         - "plugins.security.ssl.http.pemcert_filepath=certs/node.pem"

         - "plugins.security.ssl.http.pemkey_filepath=certs/node-key.pem"

         - "plugins.security.ssl.http.pemtrustedcas_filepath=certs/root-ca.pem"

         - "plugins.security.ssl.transport.pemcert_filepath=certs/node.pem"

         - "plugins.security.ssl.transport.pemkey_filepath=certs/node-key.pem"

         - "plugins.security.ssl.transport.pemtrustedcas_filepath=certs/root-ca.pem"

       ulimits:

         memlock:

           soft: -1

           hard: -1

         nofile:

           soft: 65536

           hard: 65536

       volumes:

         - opensearch-data:/usr/share/opensearch/data

         - ./certs:/usr/share/opensearch/config/certs

       ports:

         - 9200:9200

         - 9600:9600

       networks:

         - opensearch-net

   

     opensearch-dashboards:

       image: opensearchproject/opensearch-dashboards:latest

       environment:

         - 'OPENSEARCH_HOSTS=["https://opensearch:9200"]'

         - "OPENSEARCH_SSL_VERIFICATIONMODE=none"

         - "OPENSEARCH_SECURITY_ADMIN_PASSWORD=myStrongPassword123!"

         - "SERVER_SSL_ENABLED=true"

         - "SERVER_SSL_CERTIFICATE=/usr/share/opensearch-dashboards/config/certs/node.pem"

         - "SERVER_SSL_KEY=/usr/share/opensearch-dashboards/config/certs/node-key.pem"

       ports:

         - 5601:5601

       volumes:

         - ./certs:/usr/share/opensearch-dashboards/config/certs

       depends_on:

         - opensearch

       networks:

         - opensearch-net

   

   volumes:

     opensearch-data:

   

   networks:

     opensearch-net:

   

3. Start the containers:

   docker compose up -d

   

2. Installing OpenSearch MCP Server

1. Install the OpenSearch MCP server using pipx:

   pipx install opensearch-mcp-server

   

3. Configuring LLMs with MCP server configurations

1. Amazon Q: Create a mcp.json file in your Amazon Q configuration directory (~/.aws/amazonq/mcp.json)
2. Anthropic Claude: Create a claude_desktop_config.json file in your Claude Desktop configuration directory (~/Library/Application Support/Claude/claude_desktop_config.json)

   {

     "mcpServers": {

       "opensearch-mcp-server": {

         "command": "/Users/YOUR_USERNAME/.local/bin/opensearch-mcp-server",

         "args": [],

         "env": {

           "OPENSEARCH_URL": "https://localhost:9200",

           "OPENSEARCH_SSL_VERIFY": "none",

           "OPENSEARCH_USERNAME": "admin",

           "OPENSEARCH_PASSWORD": "myStrongPassword123!"

         }

       }

     }

   }

   

Replace YOUR_USERNAME with your system username or specify the location where the opensearch-mcp-server is installed.

4. Verifying the Setup

1. Check OpenSearch is running:

   curl -XGET -u admin:myStrongPassword123! https://localhost:9200

   

2. Verify OpenSearch Dashboards:
- Open https://localhost:5601 in your browser

3. Check indices:

   curl -XGET -u admin:myStrongPassword123! https://localhost:9200/_cat/indices

   

5. Using Natural Language with OpenSearch

Here are some example commands you can try in LLMs:

1. Create an index:

   Create a new index called "books" with title and author fields

   

2. Add documents:

   Add a document to the books index with title "The Great Gatsby" and author "F. Scott Fitzgerald"

   

3. Search documents:

   Search for books by F. Scott Fitzgerald

   

4. Get index information:

   Show me the mapping of the books index

   

Troubleshooting

1. If MCP server fails to initialize:
- Check if OpenSearch is running (docker ps)
- Verify credentials in mcp_config.json
- Check OpenSearch logs (docker compose logs opensearch)

2. If tools are not available:
- Restart the LLMs
- Check MCP server logs
- Verify the mcp_config.json syntax

Available MCP Tools

The OpenSearch MCP server provides several tools for interacting with OpenSearch:

- list_indices: List all indices.
- get_index: Returns information (mappings, settings, aliases) about one or more indices. Args: index: Name of the index
- create_index: Create a new index. Args: index: Name of the index body: Optional index configuration including mappings and settings
- delete_index: Delete an index. Args: index: Name of the index
- search_documents: Search for documents. Args: index: Name of the index body: Search query
- index_document: Creates or updates a document in the index. Args: index: Name of the index document: Document data id: Optional document ID
- get_document: Get a document by ID. Args: index: Name of the index id: Document ID
- delete_document: Delete a document by ID. Args: index: Name of the index id: Document ID
- delete_by_query: Deletes documents matching the provided query. Args: index: Name of the index body: Query to match documents for deletion
- get_cluster_health: Returns basic information about the health of the cluster.
- get_cluster_stats: Returns high-level overview of cluster statistics.
- list_aliases: List all aliases.
- get_alias: Get alias information for a specific index. Args: index: Name of the index
- put_alias: Create or update an alias for a specific index. Args: index: Name of the index name: Name of the alias body: Alias configuration
- delete_alias: Delete an alias for a specific index. Args: index: Name of the index name: Name of the alias
- general_api_request: Perform a general HTTP API request. Use this tool for any Elasticsearch/OpenSearch API that does not have a dedicated tool. Args: method: HTTP method (GET, POST, PUT, DELETE, etc.) path: API endpoint path params: Query parameters body: Request body

Security Notes

1. Always use strong passwords in production
2. Enable SSL in production environments
3. Regularly rotate credentials

Additional Resources

- OpenSearch Documentation
- Amazon Q Documentation
- MCP Protocol Documentation