mcpField
M
MCP Vertica

MCP Vertica

Local
github.com/nollehDatabases7Updated 51w ago

About MCP Vertica

MCP Vertica connects MCP clients to Vertica databases for query execution and schema operations. It includes result streaming, bulk loading, connection controls, and configurable operation and schema permissions.

Key features

  1. 01Execute SQL queries
  2. 02Stream query results in batches
  3. 03Bulk-load data with COPY
  4. 04Inspect tables, indexes, and views
  5. 05Manage transactions and connections
  6. 06Configure operation and schema permissions

Use cases

  1. 01Run SQL against a Vertica database
  2. 02Process large query results in batches
  3. 03Load large datasets with COPY
  4. 04Restrict write or DDL access by operation and schema

MCP support

Exposes Vertica query, data-loading, and schema-management tools to MCP clients. It can be configured for client use through uvx or Docker.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"vertica": {
"command": "uvx",
"args": [
"mcp-vertica"
]
}
}
}
Run commanduvx mcp-vertica

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
pypimcp-vertica

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

MseeP.ai Security Assessment Badge

MCP Vertica

PyPI version License: MIT Downloads smithery badge MCP Community

🏆 First implementation of Vertica MCP Server • Learn more

✅ Listed in Model Context Protocol Official Registry

A Vertica MCP(model-context-protocol) Server

Vertica MCP server
Example: MCP Server Setting

Create or edit the file your mcp client config file with the following content:

UVX
{
  "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
      }
    }
  }
}

Or with args

{
  "mcpServers": {
    "vertica": {
      "command": "uvx",
      "args": [
        "mcp-vertica",
        "--host=localhost",
        "--db-port=5433",
        "--database=VMart",
        "--user=dbadmin",
        "--password=test_password",
        "--connection-limit=10"
      ]
    }
  }
}
Docker
{
  "mcpServers": {
    "vertica": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "nolleh/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
      }
    }
  }
}

Note

  • For boolean flags like --ssl or --ssl-reject-unauthorized, simply add the flag (e.g., "--ssl") to enable it, or omit it to disable.
  • For an empty password, use an empty string as shown above.

Features

Database Connection Management
  • Connection pooling with configurable limits
  • SSL/TLS support
  • Automatic connection cleanup
  • Connection timeout handling
Query Operations
  • Execute SQL queries
  • Stream large query results in batches
  • Copy data operations
  • Transaction management
Schema Management
  • Table structure inspection
  • Index management
  • View management
  • Constraint information
  • Column details
Security Features
  • Operation-level permissions (INSERT, UPDATE, DELETE, DDL)
  • Schema-specific permissions
  • SSL/TLS support
  • Password masking in logs

Tools

Database Operations
  1. execute_query

    • Execute SQL queries
    • Support for all SQL operations
  2. stream_query

    • Stream large query results in batches
    • Configurable batch size
  3. copy_data

    • Bulk data loading using COPY command
    • Efficient for large datasets
Schema Management
  1. get_table_structure

    • Get detailed table structure
    • Column information
    • Constraints
  2. list_indexes

    • List all indexes for a table
    • Index type and uniqueness
    • Column information
  3. list_views

    • List all views in a schema
    • View definitions

Configuration

Environment Variables
VERTICA_HOST=localhost
VERTICA_PORT=5433
VERTICA_DATABASE=VMart
VERTICA_USER=newdbadmin
VERTICA_PASSWORD=vertica
VERTICA_CONNECTION_LIMIT=10
VERTICA_SSL=false
VERTICA_SSL_REJECT_UNAUTHORIZED=true
Operation Permissions
ALLOW_INSERT_OPERATION=false
ALLOW_UPDATE_OPERATION=false
ALLOW_DELETE_OPERATION=false
ALLOW_DDL_OPERATION=false
Schema Permissions
SCHEMA_INSERT_PERMISSIONS=schema1:true,schema2:false
SCHEMA_UPDATE_PERMISSIONS=schema1:true,schema2:false
SCHEMA_DELETE_PERMISSIONS=schema1:true,schema2:false
SCHEMA_DDL_PERMISSIONS=schema1:true,schema2:false

Installation

Installing via Smithery

To install Vertica Database Connector for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @nolleh/mcp-vertica --client claude
Installing Manually

Open your favorite mcp client's config file, then configure with uvx mcp-vertica

Example: Mcp Server Setting

Development

Debug Mode

When running with Docker, you can enable debug logging by setting the DEBUG environment variable:

# Run with maximum verbosity (-vvv)
docker run -e DEBUG=3 -e VERTICA_HOST=localhost ... nolleh/mcp-vertica:latest

# Run with medium verbosity (-vv)
docker run -e DEBUG=2 -e VERTICA_HOST=localhost ... nolleh/mcp-vertica:latest

# Pass additional arguments
docker run -e EXTRA_ARGS="--connection-limit=20" -e VERTICA_HOST=localhost ... nolleh/mcp-vertica:latest

In docker-compose.yml:

environment:
  DEBUG: 3  # 0=none, 1=-v, 2=-vv, 3=-vvv
  EXTRA_ARGS: "--connection-limit=20"  # Optional additional arguments
Appendix: For Testing, VerticaDB Docker Compose Example
version: "3.8"

services:
  vertica:
    # image: vertica/vertica-ce:11.1.0-0
    image: vertica/vertica-ce:latest
    platform: linux/amd64
    container_name: vertica-ce
    environment:
      VERTICA_MEMDEBUG: 2
    ports:
      - "5433:5433"
      - "5444:5444"
    volumes:
      - vertica_data:/home/dbadmin/VMart
    healthcheck:
      test:
        [
          "CMD",
          "/opt/vertica/bin/vsql",
          "-h",
          "localhost",
          "-d",
          "VMart",
          "-U",
          "dbadmin",
          "-c",
          "SELECT 1",
        ]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 30s
    restart: unless-stopped

  mcp-vertica:
    image: nolleh/mcp-vertica:latest
    container_name: mcp-vertica
    ports:
      - "8081:8081"
    environment:
      # Transport mode
      TRANSPORT: http
      PORT: 8081
      # Debug settings (0=none, 1=-v, 2=-vv, 3=-vvv)
      DEBUG: 3  # Set to 3 for maximum verbosity
      # Extra command line arguments (optional)
      # EXTRA_ARGS: "--some-flag"
      # Vertica connection settings
      VERTICA_HOST: vertica
      VERTICA_PORT: 5433
      VERTICA_DATABASE: VMart
      VERTICA_USER: dbadmin
      VERTICA_PASSWORD: ""
      VERTICA_CONNECTION_LIMIT: 10
      VERTICA_SSL: "false"
    depends_on:
      vertica:
        condition: service_healthy

volumes:
  vertica_data:
    driver: local

Then run server by following instruction Example: Mcp Server Setting, Then see everything works as fine

License

This project is licensed under the MIT License - see the LICENSE file for details.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.