📌Overview

Purpose: To provide a high-performance MCP server for storing and searching data in VikingDB, enabling efficient data management and retrieval.

Overview: VikingDB is developed by ByteDance as a robust vector database solution designed for high-speed data handling and querying. The MCP server facilitates seamless integration with VikingDB through various tools that simplify data operations.

Key Features:

vikingdb-collection-intro: Introduces users to the collections in VikingDB, helping them understand the data organization.
vikingdb-index-intro: Provides insights into the indexing structure of VikingDB, which supports fast data retrieval.
vikingdb-upsert-information: Allows users to upsert data into VikingDB, ensuring that data updates are efficient and effective.
vikingdb-search-information: Enables users to search for specific information within VikingDB, enhancing data accessibility and usability.

VikingDB MCP Server

An MCP server for VikingDB store and search.

What is VikingDB

VikingDB is a high-performance vector database developed by ByteDance.

You can easily use it following the official documentation.

Tools

The server implements the following tools:

vikingdb-collection-intro: Introduce the collection of VikingDB
vikingdb-index-intro: Introduce the index of VikingDB
vikingdb-upsert-information: Upsert information to VikingDB for later use
vikingdb-search-information: Search for information in VikingDB

Configuration

vikingdb_host: The host to use for the VikingDB server.
vikingdb_region: The region to use for the VikingDB server.
vikingdb_ak: The Access Key to use for the VikingDB server.
vikingdb_sk: The Secret Key to use for the VikingDB server.
collection_name: The name of the collection to use.
index_name: The name of the index to use.

Quickstart

Install via Smithery

To install VikingDB MCP server for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install mcp-server-vikingdb --client claude

Claude Desktop Configuration

On MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
On Windows: %APPDATA%/Claude/claude_desktop_config.json

Development/Unpublished Servers Configuration

{
  "mcpServers": {
    "mcp-server-vikingdb": {
      "command": "uv",
      "args": [
        "--directory",
        "dir to mcp-server-vikingdb",
        "run",
        "mcp-server-vikingdb",
        "--vikingdb-host", 
        "your host",
        "--vikingdb-region", 
        "your region",
        "--vikingdb-ak", 
        "your access key",
        "--vikingdb-sk", 
        "your secret key",
        "--collection-name",
        "your collection name",
        "--index-name",
        "your index name"
      ]
    }
  }
}

Published Servers Configuration

{
  "mcpServers": {
    "mcp-server-vikingdb": {
      "command": "uvx",
      "args": [
        "mcp-server-vikingdb",
        "--vikingdb-host", 
        "your host",
        "--vikingdb-region", 
        "your region",
        "--vikingdb-ak", 
        "your access key",
        "--vikingdb-sk", 
        "your secret key",
        "--collection-name",
        "your collection name",
        "--index-name",
        "your index name"
      ]
    }
  }
}

Development

Building and Publishing

Sync dependencies and update the lockfile:

uv sync

Build package distributions:

uv build

This will create source and wheel distributions in the dist/ directory.

Publish to PyPI:

uv publish

Note: You'll need to set PyPI credentials via environment variables or command flags:

Token: --token or UV_PUBLISH_TOKEN
Or username/password: --username/UV_PUBLISH_USERNAME and --password/UV_PUBLISH_PASSWORD

Debugging

Since MCP servers run over stdio, debugging can be challenging. For the best debugging experience, we strongly recommend using the MCP Inspector.

You can launch the MCP Inspector via npm with this command:

npx @modelcontextprotocol/inspector uv --directory dir_to_mcp_server_vikingdb run mcp-server-vikingdb --vikingdb-host your_host --vikingdb-region your_region --vikingdb-ak your_access_key --vikingdb-sk your_secret_key --collection-name your_collection_name --index-name your_index_name

Upon launching, the Inspector will display a URL that you can access in your browser to begin debugging.

vikingdb-mcp-server