Meilisearch MCP Server

What is Meilisearch MCP Server?

The Meilisearch MCP Server is a Model Context Protocol server that allows any MCP-compatible client (including Claude, OpenAI agents, and other LLMs) to interact with Meilisearch through natural conversation. It provides a stdio-based interface for AI assistants to manage search indices, perform searches, and handle data without requiring knowledge of the Meilisearch API.

How to use Meilisearch MCP Server?

Install the package via pip (pip install meilisearch-mcp) or uvx (uvx -n meilisearch-mcp). Configure an MCP client (e.g., Claude Desktop) by adding a server entry to claude_desktop_config.json with the command uvx and args ["-n", "meilisearch-mcp"]. Ensure a running Meilisearch instance is accessible (e.g., via Docker: docker run -d -p 7700:7700 getmeili/meilisearch:v1.28). Environment variables MEILI_HTTP_ADDR and MEILI_MASTER_KEY can set default connection settings.

Key features of Meilisearch MCP Server

• Index and document management (create, update, delete)
• Smart search with filters, sorting, and facets
• Settings configuration for ranking and relevancy
• Task monitoring and system operations tracking
• API key management for access control
• Health monitoring and system statistics

Use cases of Meilisearch MCP Server

• Create and manage search indices through natural language commands
• Add or update documents in an index via conversation
• Search across single or multiple indices with advanced filtering
• Fine-tune search settings and ranking rules using AI assistants

FAQ from Meilisearch MCP Server

What is the Meilisearch MCP Server and what alternatives exist?

The Meilisearch MCP Server connects any MCP-compatible LLM to Meilisearch for natural language interaction. A TypeScript version is also available (see README).

What are the prerequisites and runtime requirements?

Python ≥ 3.9, a running Meilisearch instance, and an MCP-compatible client (e.g., Claude Desktop, OpenAI agents).

How does data flow and where is it stored?

The server communicates with Meilisearch via its HTTP API. Data resides in the Meilisearch instance; this server does not store data locally.

What transport and authentication does it use?

It uses stdio transport. Authentication is via an API key (master key) set in environment variables or updated dynamically through conversation.

Are there any known security considerations?

While hosts and API keys can be added in chat for convenience, this is intended for development only and does not follow best MCP security practices; proper safeguards are required in production.