mcp-vertica

The MCP Vertica server exposes a set of capabilities to interact with a Vertica database through the Model Context Protocol. It provides tools to execute SQL queries, stream large result sets, copy data in bulk, and inspect or manage schema objects. Key operations include execute_query for running standard SQL, stream_query for batched result delivery, and copy_data for high-volume data loads. For schema work, you can retrieve detailed table structures with get_table_structure, enumerate indexes with list_indexes, and view existing definitions via list_views. Access is controlled by environment-based permissions to limit operations such as INSERT, UPDATE, DELETE, and DDL, and SSL/TLS can be enabled or disabled via environment flags. To run the server, configure the mcp client with uvx using the provided mcp-vertica package, and supply the Vertica connection details via environment variables or command-line arguments as shown in the example configuration.

Prerequisites:

• Python environment with uvx (or use a compatible MCP client that can run UVX commands).
• Access to a running Vertica instance (host, port, database, user, and password).

Install the MCP Vertica server via UVX:

• Ensure uvx is installed and available in your PATH.
• Install the mcp-vertica package if required by your environment (per the project's distribution, e.g., via PyPI).

Configuration steps:

1. Add the MCP server configuration to your MCP client config file (e.g., mcp-servers.json):

{
  "mcpServers": {
    "vertica": {
      "command": "uvx",
      "args": ["mcp-vertica"],
      "env": {
        "VERTICA_HOST": "localhost",
        "VERTICA_PORT": "5433",
        "VERTICA_DATABASE": "VMart",
        "VERTICA_USER": "dbadmin",
        "VERTICA_PASSWORD": "test_password",
        "VERTICA_CONNECTION_LIMIT": "10",
        "VERTICA_SSL": "false",
        "VERTICA_SSL_REJECT_UNAUTHORIZED": "true"
      }
    }
  }
}

2. Or run via Docker (optional alternative):

docker run -i --rm -e VERTICA_HOST=localhost -e VERTICA_PORT=5433 -e VERTICA_DATABASE=VMart -e VERTICA_USER=dbadmin -e VERTICA_PASSWORD=test_password -e VERTICA_CONNECTION_LIMIT=10 -e VERTICA_SSL=false -e VERTICA_SSL_REJECT_UNAUTHORIZED=true nolleh/mcp-vertica:latest uvx mcp-vertica

3. Start the server and verify it is reachable by your MCP client using the vertica configuration.

Notes:

• Replace the placeholder credentials with secure values for production.
• If you prefer Smithery, you can also install the MCP Vertica server via npx as documented in the project notes.

Environment variables control database connectivity and access. Ensure VERTICA_PASSWORD is secured and not logged. Boolean flags are represented as strings in the config (e.g., "true"/"false"). If you need to enable SSL, set VERTICA_SSL to true and provide appropriate certificates on the Vertica side. For debugging, you can pass DEBUG levels via the container environment (e.g., DEBUG=3 for verbose output). If you encounter connection issues, verify network routing between your MCP client and the Vertica host/port and confirm that the Vertica database is accepting connections from the client host.