mcp-server-mariadb-vector

@DavidRamosSal

About mcp-server-mariadb-vector

MCP server for MariaDB

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "mcp-server-mariadb-vector": {
      "command": "uv",
      "args": [
        "run",
        "--dir",
        "path/to/mcp-server-mariadb-vector/",
        "--env-file",
        "path/to/mcp-server-mariadb-vector/.env",
        "mcp_server_mariadb_vector"
      ]
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is mcp-server-mariadb-vector?

The mcp-server-mariadb-vector server provides tools that LLM agents can use to interact with a MariaDB database with vector support. It gives users a natural language interface to store and query data, and is compatible with any Model Context Protocol (MCP) client, such as Claude Desktop, Cursor, and Windsurf.

How to use mcp-server-mariadb-vector?

You need a running MariaDB instance (version 11.7 or higher) with vector support. Configure environment variables for database connection, embedding provider, and API key. Then run the server either as a Python package with uv or as a Docker container. Finally, add the server to your MCP client’s configuration file (e.g., claude_desktop_config.json) using the provided uv run command or SSE URL.

Key features of mcp-server-mariadb-vector

Create, delete, and list vector stores in MariaDB.
Add documents with optional metadata to a vector store.
Query a vector store using semantic search.
Uses OpenAI’s embedding models (e.g., text-embedding-3-small).
Works with any MCP-compatible client or framework.

Use cases of mcp-server-mariadb-vector

Provide contextual knowledge base to LLM agent conversations.
Store and retrieve conversation history with LLM agents.
Enable semantic search over personal or domain‑specific document collections.

FAQ from mcp-server-mariadb-vector

What MariaDB version is required?

MariaDB 11.7 or higher with vector support is required.

How can I run the server?

You can run the server using uv (Python package) or as a Docker container. Both methods are described in the README.

Which embedding provider is supported?

Currently, the server supports OpenAI embedding models. The default provider is openai and the default model is text-embedding-3-small.

How do I connect the server to Claude Desktop, Cursor, or Windsurf?

Add an entry for the server to the client’s MCP configuration file (e.g., claude_desktop_config.json) with the uv run command and a .env file, or use the SSE URL http://localhost:8000/sse if running as a Docker container.

What environment variables are required?

Required variables include MARIADB_HOST, MARIADB_PORT, MARIADB_USER, MARIADB_PASSWORD, MARIADB_DATABASE, EMBEDDING_PROVIDER, EMBEDDING_MODEL, and OPENAI_API_KEY. Defaults are provided for host (127.0.0.1), port (3306), provider (openai), and model (text-embedding-3-small).

Frequently asked questions

What MariaDB version is required?

MariaDB 11.7 or higher with vector support is required.

How can I run the server?

You can run the server using `uv` (Python package) or as a Docker container. Both methods are described in the README.

Which embedding provider is supported?

Currently, the server supports OpenAI embedding models. The default provider is `openai` and the default model is `text-embedding-3-small`.

How do I connect the server to Claude Desktop, Cursor, or Windsurf?

Add an entry for the server to the client’s MCP configuration file (e.g., `claude_desktop_config.json`) with the `uv run` command and a `.env` file, or use the SSE URL `http://localhost:8000/sse` if running as a Docker container.

What environment variables are required?

Required variables include `MARIADB_HOST`, `MARIADB_PORT`, `MARIADB_USER`, `MARIADB_PASSWORD`, `MARIADB_DATABASE`, `EMBEDDING_PROVIDER`, `EMBEDDING_MODEL`, and `OPENAI_API_KEY`. Defaults are provided for host (127.0.0.1), port (3306), provider (openai), and model (text-embedding-3-small).