mcp

Overview

The Yandex Cloud MCP Servers project provides a comprehensive set of Model Context Protocol servers that enable AI assistants to interact with various Yandex Cloud services. The servers are organized into categories including deployment tools, search and knowledge resources, serverless computing services, and data platform utilities. Each server implements the MCP protocol to expose specific cloud management capabilities through standardized tool interfaces, allowing users to manage infrastructure through conversational AI interfaces.

Try asking AI

After installing, here are 7 things you can ask your AI assistant:

• create_function

  Create a new serverless function in Yandex Cloud

  Try:Create a new function that processes image uploadsTry:Create a function named 'process-orders' with a 128MB memory limit
• deploy_function

  Deploy a serverless function to Yandex Cloud

  Try:Deploy my local function to the 'production' versionTry:Deploy function with new tags for testing
• create_container

  Deploy a new serverless container in Yandex Cloud

  Try:Create a new container from my Docker imageTry:Create a container with 2 vCPUs and 4GB RAM
• create_trigger

  Create an event-driven trigger for serverless functions or containers

  Try:Create a trigger that runs my function when new files are uploaded to storageTry:Create a timer trigger that runs every hour
• create_workflow

  Create a serverless workflow with YAML specifications

  Try:Create a workflow to process user registrationsTry:Create a scheduled workflow that runs daily reports
• search_documentation

  Search Yandex Cloud documentation in real-time

  Try:How do I set up a VPC in Yandex Cloud?Try:Find documentation for YDB managed service
• web_search

  Search the web using Yandex Search

  Try:Find information about latest AI trendsTry:Search for best practices in cloud security
• create_deployment

  Deploy a simple application with Compute, VPC, IAM, and Storage

  Try:Deploy a web server with Nginx on VMTry:Create a deployment with load balancer and auto-scaling
• search_tables

  Search for tables and views in the data catalog

  Try:Find all tables related to user profilesTry:Search for views with financial data
• create_api_gateway

  Create and manage API gateways with OpenAPI specs

  Try:Create an API gateway for my mobile app backendTry:Set up API gateway with custom domain
• create_mcp_gateway

  Configure MCP gateways that invoke functions, containers, and workflows

  Try:Create a gateway that exposes my function as an HTTP endpointTry:Set up gateway to trigger workflow via external call
• create_serverless_container

  Deploy a containerized application with revision management

  Try:Deploy my container app with auto-scalingTry:Create a new revision of my container

Note: Inferred from server descriptions as the actual tool names are not explicitly listed in the main README. Each server's specific documentation would have the exact tool names and signatures.

Comparable tools

Installation

{
  "mcpServers": {
    "yandex-cloud-toolkit": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y", "@yandex-cloud/mcp",
        "-s", "toolkit"
      ]
    }
  }
}

{
  "mcpServers": {
    "yandex-cloud-toolkit": {
      "type": "streamableHttp",
      "url": "https://toolkit.mcp.cloud.yandex.net/mcp",
      "headers": {
        "Authorization": "Bearer <YC IAM Token>"
      }
    }
  }
}

FAQ

How do I authenticate with Yandex Cloud MCP servers?

You can authenticate either via OAuth (browser-based) using the NPM client or by using an IAM token with the Streamable HTTP method. The OAuth method is recommended and doesn't require additional installation.

Do I need to install Yandex Cloud CLI to use these servers?

No, the CLI is optional. It's only required if you choose to authenticate using the CLI method or if you need to obtain an IAM token for the HTTP authentication method.

How many MCP servers are included in this project?

There are currently 10 specialized MCP servers covering deployment, search & knowledge, serverless (6 different services), and data platform categories.

Compare mcp with