Yandex Cloud MCP Servers

Specialized MCP servers for interacting with Yandex Cloud using the MCP protocol.

Table of Contents

• Yandex Cloud MCP Servers
  • Table of Contents
  • About Model Context Protocol (MCP)
  • YC MCP Servers Concept
  • Installing MCP Servers
    • Available MCP Servers
      • Deployment
      • Search & Knowledge
      • Serverless
      • Data Platform
    • Configuration
      • NPM Client (Stdio)
      • Streamable HTTP
  • License

About Model Context Protocol (MCP)

MCP (Model Context Protocol) is an open-source standard for connecting AI applications to external systems.

Using MCP, AI applications like Claude or ChatGPT can connect to data sources (e.g. local files, databases), tools (e.g. search engines, calculators) and workflows (e.g. specialized prompts) — enabling them to access key information and perform tasks.

— Model Context Protocol documentation (Anthropic)

MCP solves a fundamental challenge in AI development: connecting AI assistants to the systems and data they need to be useful. Without MCP, developers must build custom integrations for each AI platform and data source combination. This creates duplicated work and maintenance burden.

MCP provides a universal standard that works across different AI applications. This dramatically reduces development time and makes AI integrations more reliable and maintainable.

YC MCP Servers Concept

As the adoption of AI continues to grow across all areas of development, we sincerely believe that managing cloud infrastructure with MCP servers is part of the future of cloud management.

MCP servers of cloud providers also help to open the door to the world of cloud technologies for non-tech people, e.g. product manager who wants to validate their new idea but doesn't have sufficient skills - it's never been easier to build and deploy the "proof-of-concept" of any service than by using AI + MCP.

Using provided MCP servers, you can manage Yandex Cloud infrastructure either for "VibeDevOps" and complex AI-automation scenarios, which extremely simplifies working with the cloud infrastructure.

Installing MCP Servers

Available MCP Servers

Deployment

Search & Knowledge

Serverless

Data Platform

Configuration

To connect your assistant with MCP servers, you have to update your assistant's configuration (e.g. Cline, Roo Code or Claude Desktop) by adding chosen server.

Most MCP servers need to be authorized in Yandex Cloud. There are several ways to configure and authorize:

1. NPM Client (Stdio)

Provides automatic authentication by integrating with Yandex Cloud CLI.

Prerequisites:

• Server-specific Yandex Cloud roles.
• Node.js 18.0.0 or higher
• Yandex Cloud CLI (yc) installed with configured user profile

See the package documentation for more details.

Configuration example:

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

Here toolkit stands for the server name, e.g. search, functions, docs. You can find the particular server's name in its npm client configuration section, e.g. this one for Yandex Search MCP Server.

2. Streamable HTTP

Prerequisites:

• Server-specific Yandex Cloud roles.

• IAM token. You can get it using Yandex Cloud CLI:

  • yc iam create-token for user account
  • yc iam create-token --impersonate-service-account-id <service-account-id> for service account

  The IAM token has a maximum lifespan of 12 hours. After expiration, it must be rotated.

Configuration example:

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

You can find the particular server's url in its streamable http configuration section, e.g. this one for Yandex Search MCP Server.

License

This project is licensed under the Apache-2.0 License.