Meilisearch MCP Server

Overview

Meilisearch MCP allows you to interact with Meilisearch, a powerful open-source search engine, directly through LLM interfaces. This integration enables you to manage your search infrastructure, perform queries, and handle administrative tasks without switching contexts.

Installation

Prerequisites

• Python 3.9 or higher
• A running Meilisearch instance
• Node.js (for testing with MCP Inspector)

Installation Steps

1. Install the package using pip:

pip install meilisearch-mcp

2. Configure your environment variables (optional):

MEILI_HTTP_ADDR=http://localhost:7700  # Default Meilisearch URL
MEILI_MASTER_KEY=your_master_key       # Optional: Default Meilisearch API key

Usage

Running the Server

Start the MCP server by running:

python -m meilisearch_mcp

Or with uvx:

uvx -n meilisearch-mcp

Connecting to Meilisearch

The MCP server automatically connects to the Meilisearch instance specified in your environment variables. You can view or update these connection settings at runtime using the provided tools:

• Use get-connection-settings to view current connection details
• Use update-connection-settings to change the URL or API key

Working with Indices

You can create, list, and manage indices:

1. Create a new index:

Use the create-index tool with an indexUid and optional primaryKey

2. List all available indices:

Use the list-indexes tool to see all indices in your Meilisearch instance

3. Get detailed metrics for a specific index:

Use get-index-metrics with an indexUid to see statistics about the index

Document Management

Add and retrieve documents from your indices:

1. Add documents:

Use add-documents with indexUid and documents (JSON array) parameters

2. Retrieve documents:

Use get-documents with indexUid and optional pagination parameters

Searching

The search functionality is flexible and powerful:

1. Search in a specific index:

Use search with query and indexUid parameters

2. Search across all indices:

Use search with just the query parameter

3. Advanced search with filtering and sorting:

Add filter, sort, limit, and offset parameters to refine your search

Settings Management

Configure how your search engine behaves:

1. View current settings:

Use get-settings with an indexUid parameter

2. Update settings:

Use update-settings with indexUid and the settings you want to change

Task and System Monitoring

Monitor the status of operations and system health:

1. Track tasks:

Use get-tasks with optional filters to see the status of operations

2. Check system health:

Use health-check or get-health-status to verify your Meilisearch instance is running properly

Troubleshooting

If you encounter connection issues:

1. Verify your Meilisearch instance is running
2. Check that the URL and API key are correct using get-connection-settings
3. Update connection settings if needed using update-connection-settings

For task failures:

1. Use get-task to check the detailed status of a specific task
2. Review error messages for clues about what went wrong